A muscle function assessment system for residual limbs of forearm amputees based on electromyographic signals
Through an evaluation system based on electromyography signals, the problem of irregularity and poor reliability in the evaluation of residual limbs in patients with forearm amputation was solved, and the multi-angle and multi-level evaluation of muscle function was achieved, which improved the accuracy and comprehensiveness of the evaluation, and for recovery. Training provides better guidance.
Patent Information
- Application Number
- CN202410196862.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-02-22
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-02-22
AI Technical Summary
The prior art has problems of irregularity, poor reliability and lack of comprehensive analysis of residual muscle function evaluation in patients with forearm amputation, resulting in inaccurate and incomplete evaluation results.
Using an evaluation system based on electromyography signals, including signal acquisition and transmission module, signal processing module and signal analysis module, the electromyography signal of residual limbs is collected and analyzed through surface electromyography technology, muscle activity analysis, muscle antagonism analysis and characteristics Similarity analysis and generate detailed evaluation reports.
It improves the accuracy and comprehensiveness of the evaluation of residual muscle function, provides a reliable evaluation method, helping rehabilitation physicians to more accurately guide prosthetic assembly and rehabilitation training, and promotes the development of rehabilitation technology for patients with forearm amputation.
Smart Images

Figure CN118078312B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of rehabilitation detection, and in particular to a muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals. Background Art
[0002] After losing hand function, patients with forearm amputation often need prostheses to resume functional activities in daily life. Currently, most commercial myoelectric prostheses control prosthetic movements or switch different degrees of freedom through different movement patterns of residual limb muscles. Accurately assessing the functional state of residual limb muscles to adapt the prosthesis and conduct targeted training is crucial for patients to use myoelectric prostheses.
[0003] Rehabilitation physicians usually evaluate the patient's residual limb before prosthesis fitting and during training. However, traditional evaluation methods often use manual muscle testing (MMT) and clinical observation to assess residual limb muscle strength, skin condition, shape, circumference and other parameters, without analysis and evaluation of residual limb electromyographic signals. This results in the current residual limb assessment being unable to provide comprehensive data to achieve optimal prosthesis fitting and effectively guide the rehabilitation training process. An improved method is urgently needed to accurately evaluate the functional status of residual limb muscles.
[0004] However, the existing residual limb muscle function assessment for forearm amputees has the following problems:
[0005] 1) Non-standard: When assessing residual limbs, rehabilitation therapists and assessors currently use verbal language to guide patients to perform the movements of the missing hand, which is not vivid and standard. In the noisy environment of the rehabilitation department, patients may not accurately receive the rehabilitation therapist's guidance information.
[0006] 2) Poor reliability: Traditional assessment methods, such as manual strength testing and clinical observation, rely on the experience and judgment of rehabilitation physicians and are highly subjective. They may be affected by differences in the experience of rehabilitation physicians, resulting in poor reliability of the assessment results.
[0007] 3) Lack of comprehensive analysis of electromyographic data: The traditional evaluation method of manual muscle strength test makes a rough five-level diagnosis of residual limb muscle strength. It does not conduct multi-channel electromyographic acquisition and analysis of the residual limb during different movement tasks. As a result, the rehabilitation physician's understanding of the residual limb muscle function is partial, non-holistic, and comprehensive. Summary of the invention
[0008] The purpose of the present invention is to provide a muscle function evaluation system for the residual limb of a forearm amputee based on electromyographic signals, comprising: a signal acquisition and transmission module, a signal processing module, and a signal analysis module.
[0009] The signal acquisition and transmission module is used to acquire residual limb electromyographic signals of forearm amputation patients in different movements, and transmit the residual limb electromyographic signals to the signal processing module.
[0010] The signal processing module is used to preprocess and extract features of the collected residual limb electromyographic signals.
[0011] The signal analysis module is used to perform muscle activity analysis, muscle antagonism analysis and feature similarity analysis on the extracted features, and generate an evaluation report on the residual limb muscle function of forearm amputees.
[0012] Furthermore, the signal acquisition and transmission module may acquire electromyographic signals of the residual limb from a portion of the muscle including but not limited to the flexor carpi ulnaris, flexor carpi radialis, extensor carpi ulnaris, and extensor carpi radialis longus.
[0013] Furthermore, the muscle function assessment system also includes a gesture guidance module.
[0014] The gesture guidance module guides the patient to use the residual limb to complete the gesture action in the form of images and / or text instructions through the display, and the playback time of each gesture action is t 0c c represents the cth gesture. The interval between each two gestures is t 0u .
[0015] The gesture actions include but are not limited to resting gesture actions, single-degree-of-freedom movements, and multi-degree-of-freedom movements.
[0016] The single degree of freedom movement includes wrist extension, wrist flexion, wrist internal rotation, and wrist external rotation.
[0017] The multi-degree-of-freedom movements include making a fist, pinching with five fingers, and extending the palm.
[0018] When the gesture action is a resting gesture action or a single-degree-of-freedom motion, the residual limb electromyographic signal is a single-degree-of-freedom motion electromyographic signal.
[0019] When the gesture action is a multi-degree-of-freedom movement, the residual limb electromyographic signal is a multi-degree-of-freedom movement electromyographic signal, including electromyographic signals from at least two flexor muscles and / or extensor muscles of the residual limb.
[0020] Furthermore, the preprocessing includes filtering and data segmentation.
[0021] The data segmentation is used to select a window for feature extraction.
[0022] Furthermore, the features include but are not limited to root mean square value, integrated electromyographic value, frequency domain waveform length, absolute mean value of difference, peak mean, average rectified electromyographic value, absolute standard deviation value of difference, maximum fractal length, time domain waveform length, power spectral density, phase lock value, standard deviation, peak frequency, and median frequency.
[0023] Further, the calculation formula of the root mean square value RMS is as follows:
[0024]
[0025] Where i is the frequency point number, i = 1, 2, ..., N, N is the total number of frequency points. i is the electromyographic signal value at the i-th frequency point.
[0026] The calculation formula of the integrated electromyographic value IEMG is as follows:
[0027]
[0028] The calculation formula of the frequency domain waveform length FDWL is as follows:
[0029]
[0030] Where PSD(i) and PSD(i+1) are the power spectral densities of the electromyographic signal at the i-th and i+1-th frequency points, respectively, and △f is the frequency interval between adjacent frequency points.
[0031] The calculation formula of the difference absolute mean value DAMV is as follows:
[0032]
[0033] In the formula, x i+1 is the electromyographic signal value at the i+1th frequency point.
[0034] The calculation formula of the peak mean value MPV is as follows:
[0035]
[0036] Where j is the peak number, j = 1, 2, ..., m, and m is the total number of peaks. j is the jth peak in the electromyographic signal.
[0037] The calculation formula of the average rectified electromyographic value MRV is as follows:
[0038]
[0039] The calculation formula of the differential absolute standard deviation value DASDV is as follows:
[0040]
[0041] The calculation formula of the maximum fractal length MFL is as follows:
[0042] MFL=max(|x i+1 -x i |) (8)
[0043] The calculation formula of the time domain waveform length TWL is as follows:
[0044]
[0045] The power spectrum density is calculated by performing Fourier transform on the electromyographic signal in the frequency domain.
[0046] The phase locking value is obtained by calculation in the frequency domain.
[0047] The calculation formula of the standard deviation STD is as follows:
[0048]
[0049] In the formula, is the mean value of the electromyographic signal.
[0050] The peak frequency is determined by the peak in the frequency spectrum of the electromyographic signal.
[0051] The median frequency is determined by the median value in the frequency spectrum of the electromyographic signal.
[0052] Furthermore, the muscle activity analysis method is to visualize the extracted features, display the activity intensity of different muscles in different actions and compare the activation levels of different muscles, so as to obtain the electromyographic level analysis of different actions.
[0053] The method for visualizing the extracted features includes but is not limited to comparing the RMS histograms of the surface electromyography of different muscles under a single action, comparing the surface electromyography IEMG histograms of muscles under different actions, comparing the radar charts of the surface electromyography FWL of different muscles under a single action, and comparing the radar charts of the surface electromyography RMS of different muscles under a single action.
[0054] Furthermore, the muscle antagonism analysis is to obtain an evaluation result of the antagonism between the agonist muscle and the antagonist muscle during wrist extension and wrist flexion by calculating the antagonism coefficients of the radial side and the ulnar side.
[0055] The antagonism coefficient includes a wrist extension antagonism coefficient and a wrist flexion antagonism coefficient.
[0056] The calculation formulas of the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are as follows:
[0057]
[0058]
[0059] Where, RMS 屈肌 It is the sum of the root mean square values of the electromyographic signals of the radial flexor carpi and the ulnar flexor carpi in patients with forearm amputation. RMS 伸肌 It is the sum of the RMS values of the EMG signals of the radial extensor carpi and the ulnar extensor carpi in patients with forearm amputation.
[0060] Further, the step of performing feature similarity analysis on the extracted features includes:
[0061] 1) Standardize the extracted feature sequences and scale each standardized feature sequence using the L2 Euclidean norm. The scaled standardized feature sequence τ n As shown below:
[0062]
[0063] Where n is the sequence number of the electromyographic feature, ||T n || 2 represents the L2 Euclidean norm. mean(T n ) is the electromyographic characteristic sequence T n The mean of .
[0064] Among them, the nth electromyographic feature sequence T n As shown below:
[0065] T n =[t1, t2, ..., t g ] (14)
[0066] In the formula, t k is the electromyographic feature corresponding to the kth action, k = 1, 2, ..., g, where g is the total number of actions.
[0067] 2) All scaled and standardized feature sequences τ n Combined into a set of electromyographic signal feature sequences F corresponding to different muscle positions p , as shown below:
[0068] F p =[τ1,τ2,...,τ d ] (15)
[0069] Where d is the total number of myoelectric feature sequences and p represents the muscle position number of the forearm.
[0070] 3) The Pearson correlation coefficient was used to analyze the correlation between the characteristic sequences of myoelectric signals at different muscle positions in the forearm of forearm amputees and healthy subjects, as shown below:
[0071]
[0072] Where, P is the correlation coefficient between the EMG signal characteristics of the residual limb of the amputee and the EMG signal characteristics of the healthy subjects, P∈[-1,1]. p=1,2,...,q, q is the total number of muscle positions in the forearm. Represents the mean values of the characteristic sequences of electromyographic signals of different muscle positions in the forearm of patients with forearm amputation and healthy subjects. p Represents the characteristic sequence of electromyographic signals at the pth muscle position in the forearm of a healthy subject.
[0073] 4) According to the similarity of the electromyographic signal characteristics of the residual limbs of amputees, the muscle function evaluation index is obtained as follows:
[0074]
[0075] Furthermore, the evaluation report includes a report summary, experimental results, data analysis, and a summary.
[0076] The report summary includes the patient's name, age, gender, residual limb condition, and duration of amputation.
[0077] The experimental results include the extracted characteristic values of multiple electromyographic signals, the antagonism coefficients of the extensor muscles and the flexor muscles, and the characteristic similarities.
[0078] The data analysis includes the myoelectric signals of different forearm muscles of forearm amputees completing each action, the antagonistic performance of agonist muscles and antagonist muscles, and muscle function evaluation indicators obtained by feature similarity analysis.
[0079] The summary evaluates the myoelectric prosthesis control ability of the forearm amputee patients completing each action based on the analysis results of the myoelectric signals of different forearm muscles, the antagonistic performance of the agonist and antagonist muscles, and the muscle function evaluation indexes, and provides rehabilitation training suggestions.
[0080] If either the wrist extension antagonism coefficient or the wrist flexion antagonism coefficient is less than 0.5, and the muscle function evaluation index is weak or relatively weak, it is not suitable for fitting a myoelectric prosthesis. It is recommended to undergo residual limb rehabilitation training before re-evaluation.
[0081] If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is weak or relatively weak, it is recommended to conduct residual limb rehabilitation training before fitting the myoelectric prosthesis.
[0082] If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is strong or relatively strong, it is recommended to fit a myoelectric prosthesis.
[0083] The technical effect of the present invention is unquestionable. The present invention provides a method for evaluating the residual limb muscle function of forearm amputation patients based on electromyographic signals. The method uses surface electromyography (sEMG) technology to collect and analyze the electromyographic signals of the residual limbs of forearm amputation patients in different movements, evaluates the residual limb muscle function of forearm amputation patients, and provides an effective residual limb function evaluation method for rehabilitation physicians and amputees.
[0084] The present invention performs muscle activity analysis, muscle antagonism analysis and feature similarity analysis on the collected electromyographic signals, provides multi-angle and multi-level electromyographic information, and improves the accuracy and comprehensiveness of residual limb muscle function evaluation.
[0085] The innovative method of the present invention not only proposes a highly reliable evaluation method and improves the accuracy of the evaluation, but also provides more accurate guidance data for prosthetic assembly and rehabilitation training, thereby promoting the development of rehabilitation technology for forearm amputees.
[0086] The present invention has developed a set of gesture guidance programs in the form of images, which provide intuitive action guidance for subjects in the form of images and text instructions to help subjects complete gesture actions in a specified order and time. The subjects' healthy palm and residual limb perform the same action at the same time, ensuring the effectiveness of the action.
[0087] The present invention proposes to place 8 electromyographic acquisition electrodes on the four flexor and extensor muscles on the surface of the forearm residual limb, and improve the signal-to-noise ratio of the electromyographic signal by subtracting the electromyographic signals of 2 channels, and then record the multi-channel electromyographic signals of the subject's residual limb when performing a specified gesture action task, and extract and analyze ≮9 time domain and frequency domain features to provide an important supplement to the clinical evaluation of clinical rehabilitation physicians, so as to realize an objective auxiliary evaluation of electromyographic signal characteristics in addition to manual strength testing and clinical observation.
[0088] The present invention utilizes the coordination mode of four muscles to intuitively display the change in the strength amplitude of the four muscles when the muscles are in a specific gesture action.
[0089] The present invention can not only realize the myoelectric level analysis of resting gestures and simple movements with a single degree of freedom such as wrist flexion and extension, wrist internal and external rotation, but also can realize the myoelectric level analysis of movements with multiple degrees of freedom such as fist clenching, palm extension, and five-finger pinching. BRIEF DESCRIPTION OF THE DRAWINGS
[0090] Figure 1 It is a schematic diagram of the use of the present invention; Figure 1 (a) is a schematic diagram of the overall use of the present invention; Figure 1 (b) is a schematic diagram of use on a patient's forearm;
[0091] Figure 2 It is a block diagram of the residual limb function assessment system of the present invention;
[0092] Figure 3 is a flow chart of gesture guidance of the present invention;
[0093] Figure 4 It is a flow chart of myoelectric signal processing of the present invention;
[0094] Figure 5 This is a flow chart of the electromyographic signal analysis of the present invention. DETAILED DESCRIPTION
[0095] The present invention is further described below in conjunction with the embodiments, but it should not be understood that the above subject matter of the present invention is limited to the following embodiments. Without departing from the above technical ideas of the present invention, various substitutions and changes are made according to the common technical knowledge and customary means in the art, which should all be included in the protection scope of the present invention.
[0096] Embodiment 1:
[0097] See also Figures 1 to 5 A muscle function evaluation system for the residual limb of a forearm amputee based on electromyographic signals includes: a signal acquisition and transmission module, a signal processing module, and a signal analysis module.
[0098] The signal acquisition and transmission module is used to acquire residual limb electromyographic signals of forearm amputation patients in different movements, and transmit the residual limb electromyographic signals to the signal processing module.
[0099] The signal processing module is used to preprocess and extract features of the collected residual limb electromyographic signals.
[0100] The signal analysis module is used to perform muscle activity analysis, muscle antagonism analysis and feature similarity analysis on the extracted features, and generate an evaluation report on the residual limb muscle function of forearm amputees.
[0101] Embodiment 2:
[0102] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals. The main technical content is shown in Example 1. Furthermore, the signal acquisition and transmission module collects electromyographic signals of the residual limb from the following parts, including but not limited to the ulnar flexor carpi, radial flexor carpi, ulnar extensor carpi, and radial extensor carpi longus.
[0103] Embodiment 3:
[0104] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 1 to 2. Furthermore, the muscle function assessment system also includes a gesture guidance module.
[0105] The gesture guidance module guides the patient to use the residual limb to complete the gesture action in the form of images and / or text instructions through the display, and the playback time of each gesture action is t 0c c represents the cth gesture. The interval between each two gestures is t 0u .
[0106] The gesture actions include but are not limited to resting gesture actions, single-degree-of-freedom movements, and multi-degree-of-freedom movements.
[0107] The single degree of freedom movement includes wrist extension, wrist flexion, wrist internal rotation, and wrist external rotation.
[0108] The multi-degree-of-freedom movements include making a fist, pinching with five fingers, and extending the palm.
[0109] When the gesture action is a resting gesture action or a single-degree-of-freedom motion, the residual limb electromyographic signal is a single-degree-of-freedom motion electromyographic signal.
[0110] When the gesture action is a multi-degree-of-freedom movement, the residual limb electromyographic signal is a multi-degree-of-freedom movement electromyographic signal, including electromyographic signals from at least two flexor muscles and / or extensor muscles of the residual limb.
[0111] Embodiment 4:
[0112] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 1 to 3. Furthermore, the preprocessing includes filtering and data segmentation.
[0113] The data segmentation is used to select a window for feature extraction.
[0114] Embodiment 5:
[0115] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 1 to 4. Furthermore, the characteristics include but are not limited to root mean square value, integrated electromyographic value, frequency domain waveform length, absolute average value of difference, peak mean, average rectified electromyographic value, absolute standard deviation value of difference, maximum fractal length, time domain waveform length, power spectral density, phase locking value, standard deviation, peak frequency, and median frequency.
[0116] Embodiment 6:
[0117] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Embodiments 1 to 5. Further, the calculation formula of the root mean square value RMS is as follows:
[0118]
[0119] Where i is the frequency point number, i = 1, 2, ..., N, N is the total number of frequency points. i is the electromyographic signal value at the i-th frequency point.
[0120] The calculation formula of the integrated electromyographic value IEMG is as follows:
[0121]
[0122] The calculation formula of the frequency domain waveform length FDWL is as follows:
[0123]
[0124] Where PSD(i) and PSD(i+1) are the power spectral densities of the electromyographic signal at the i-th and i+1-th frequency points, respectively, and △f is the frequency interval between adjacent frequency points.
[0125] The calculation formula of the difference absolute mean value DAMV is as follows:
[0126]
[0127] In the formula, x i+1 is the electromyographic signal value at the i+1th frequency point.
[0128] The calculation formula of the peak mean value MPV is as follows:
[0129]
[0130] Where j is the peak number, j = 1, 2, ..., m, and m is the total number of peaks. j is the jth peak in the electromyographic signal.
[0131] The calculation formula of the average rectified electromyographic value MRV is as follows:
[0132]
[0133] The calculation formula of the differential absolute standard deviation value DASDV is as follows:
[0134]
[0135] The calculation formula of the maximum fractal length MFL is as follows:
[0136] MFL=max(|x i+1 -x i |) (8)
[0137] The calculation formula of the time domain waveform length TWL is as follows:
[0138]
[0139] The power spectrum density is calculated by performing Fourier transform on the electromyographic signal in the frequency domain.
[0140] The phase locking value is obtained by calculation in the frequency domain.
[0141] The calculation formula of the standard deviation STD is as follows:
[0142]
[0143] In the formula, is the mean value of the electromyographic signal.
[0144] The peak frequency is determined by the peak in the frequency spectrum of the electromyographic signal.
[0145] The median frequency is determined by the median value in the frequency spectrum of the electromyographic signal.
[0146] Embodiment 7:
[0147] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 1 to 6. Furthermore, the muscle activity analysis method is to visualize the extracted features, display the activity intensity of different muscles in different actions and compare the activation levels of different muscles, thereby obtaining electromyographic level analysis of different actions.
[0148] The method for visualizing the extracted features includes but is not limited to comparing the RMS histograms of the surface electromyography of different muscles under a single action, comparing the surface electromyography IEMG histograms of muscles under different actions, comparing the radar charts of the surface electromyography FWL of different muscles under a single action, and comparing the radar charts of the surface electromyography RMS of different muscles under a single action.
[0149] Embodiment 8:
[0150] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 1 to 7. Furthermore, the muscle antagonism analysis is performed by calculating the antagonism coefficients of the radial side and the ulnar side, thereby obtaining an evaluation result of the antagonism between the agonist muscle and the antagonist muscle during wrist extension and wrist flexion.
[0151] The antagonism coefficient includes a wrist extension antagonism coefficient and a wrist flexion antagonism coefficient.
[0152] The calculation formulas of the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are as follows:
[0153]
[0154]
[0155] Where, RMS 屈肌 It is the sum of the root mean square values of the electromyographic signals of the radial flexor carpi and the ulnar flexor carpi in patients with forearm amputation. RMS 伸肌 It is the sum of the RMS values of the EMG signals of the radial extensor carpi and the ulnar extensor carpi in patients with forearm amputation.
[0156] Embodiment 9:
[0157] A muscle function assessment system for residual limbs of forearm amputees based on electromyographic signals, the main technical content of which is shown in any one of Embodiments 1 to 8. Further, the step of performing feature similarity analysis on the extracted features comprises:
[0158] 1) Standardize the extracted feature sequences and scale each standardized feature sequence using the L2 Euclidean norm. The scaled standardized feature sequence τ n As shown below:
[0159]
[0160] Where n is the sequence number of the electromyographic feature, ||T n || 2 represents the L2 Euclidean norm. mean(T n ) is the electromyographic characteristic sequence T n The mean of .
[0161] Among them, the nth electromyographic feature sequence T n As shown below:
[0162] T n =[t1, t2, ..., t g ] (14)
[0163] In the formula, t k is the electromyographic feature corresponding to the kth action, k = 1, 2, ..., g, where g is the total number of actions.
[0164] 2) All scaled and standardized feature sequences τ n Combined into a set of electromyographic signal feature sequences F corresponding to different muscle positions p , as shown below:
[0165] F p =[τ1,τ2,...,τ d ] (15)
[0166] Where d is the total number of myoelectric feature sequences and p represents the muscle position number of the forearm.
[0167] 3) The Pearson correlation coefficient was used to analyze the correlation between the characteristic sequences of myoelectric signals at different muscle positions in the forearm of forearm amputees and healthy subjects, as shown below:
[0168]
[0169] Where, P is the correlation coefficient between the EMG signal characteristics of the residual limb of the amputee and the EMG signal characteristics of the healthy subjects, P∈[-1,1]. p=1,2,...,q, q is the total number of muscle positions in the forearm. Represents the mean values of the characteristic sequences of electromyographic signals of different muscle positions in the forearm of patients with forearm amputation and healthy subjects. p Represents the characteristic sequence of electromyographic signals at the pth muscle position in the forearm of a healthy subject.
[0170] 4) According to the similarity of the electromyographic signal characteristics of the residual limbs of amputees, the muscle function evaluation index is obtained as follows:
[0171]
[0172] Embodiment 10:
[0173] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 1 to 9. Furthermore, the assessment report includes a report summary, experimental results, data analysis, and a summary.
[0174] The report summary includes the patient's name, age, gender, residual limb condition, and duration of amputation.
[0175] The experimental results include the extracted characteristic values of multiple electromyographic signals, the antagonism coefficients of the extensor muscles and the flexor muscles, and the characteristic similarities.
[0176] The data analysis includes the myoelectric signals of different forearm muscles of forearm amputees completing each action, the antagonistic performance of agonist muscles and antagonist muscles, and muscle function evaluation indicators obtained by feature similarity analysis.
[0177] The summary evaluates the myoelectric prosthesis control ability of the forearm amputee patients completing each action based on the analysis results of the myoelectric signals of different forearm muscles, the antagonistic performance of the agonist and antagonist muscles, and the muscle function evaluation indexes, and provides rehabilitation training suggestions.
[0178] If either the wrist extension antagonism coefficient or the wrist flexion antagonism coefficient is less than 0.5, and the muscle function evaluation index is weak or relatively weak, it is not suitable for fitting a myoelectric prosthesis. It is recommended to undergo residual limb rehabilitation training before re-evaluation.
[0179] If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is weak or relatively weak, it is recommended to conduct residual limb rehabilitation training before fitting the myoelectric prosthesis.
[0180] If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is strong or relatively strong, it is recommended to fit a myoelectric prosthesis.
[0181] Embodiment 11:
[0182] See also Figures 1 to 5 A muscle function evaluation system for the residual limb of a forearm amputee based on electromyographic signals includes: a signal acquisition and transmission module, a signal processing module, and a signal analysis module.
[0183] The signal acquisition and transmission module is used to acquire residual limb electromyographic signals of forearm amputation patients in different movements, and transmit the residual limb electromyographic signals to the signal processing module.
[0184] The signal acquisition module collects the electromyographic signals of the residual limbs of forearm amputation patients under different movements, and transmits the data to the host computer software for subsequent signal processing and analysis.
[0185] The signal acquisition module sends the original electromyographic signal to the signal processing module via wireless transmission.
[0186] The signal processing module is used to preprocess and extract features of the collected residual limb electromyographic signals.
[0187] The signal analysis module is used to perform muscle activity analysis, muscle antagonism analysis and feature similarity analysis on the extracted features, and generate an evaluation report on the residual limb muscle function of forearm amputees.
[0188] Embodiment 12:
[0189] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals. The main technical content is shown in Example 11. Furthermore, the signal acquisition module uses electrodes made of Ag / Agcl (silver / silver chloride) to collect electromyographic signals of multiple muscles of the patient's residual limb.
[0190] The signal acquisition and transmission module acquires electromyographic signals of the residual limb from the following parts, including but not limited to the flexor carpi ulnaris, flexor carpi radialis, extensor carpi ulnaris, and extensor carpi radialis longus.
[0191] Embodiment 13:
[0192] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 11 to 12. Furthermore, the signal acquisition module needs to collect electromyographic signals of the amputee in seven movement movements.
[0193] The muscle function assessment system also includes a gesture guidance module.
[0194] The gesture guidance module guides the patient to use the residual limb to complete the gesture action in the form of images and / or text instructions through the display, so that the patient can intuitively understand and imitate. The playback time of each gesture action is 5 seconds. The interval between each two gesture actions is 5 seconds.
[0195] The gesture actions include but are not limited to resting gesture actions, single-degree-of-freedom movements, and multi-degree-of-freedom movements.
[0196] The single degree of freedom movement includes wrist extension, wrist flexion, wrist internal rotation, and wrist external rotation.
[0197] The multi-degree-of-freedom movements include making a fist, pinching with five fingers, and extending the palm.
[0198] When the gesture action is a resting gesture action or a single-degree-of-freedom motion, the residual limb electromyographic signal is a single-degree-of-freedom motion electromyographic signal.
[0199] When the gesture action is a multi-degree-of-freedom movement, the residual limb electromyographic signal is a multi-degree-of-freedom movement electromyographic signal, including electromyographic signals from at least two flexor muscles and / or extensor muscles of the residual limb.
[0200] Embodiment 14:
[0201] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 11 to 13. Furthermore, the preprocessing includes filtering and data segmentation.
[0202] The data segmentation is used to select a window for feature extraction.
[0203] The signal processing module filters the original electromyographic signal through a 4th-order Butterworth bandpass filter with a passband range of 20 Hz-350 Hz to remove the high-frequency noise of the signal and the interference caused by poor contact between the electrode and the skin, and then uses a 4th-order Butterworth power frequency notch filter to filter out 50 Hz power frequency interference.
[0204] The signal processing module annotates different actions of the electromyographic signal and selects an analysis window with a window length of 200ms and a step length of 100ms to further segment the multi-channel electromyographic signal in the activity segment, and obtains a series of analysis window data as a sample set for electromyographic feature extraction.
[0205] Embodiment 15:
[0206] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 11 to 14. Further, the features include but are not limited to root mean square (RMS), integrated electromyography (IEMG), frequency domain waveform length (FDWL), deviation absolute mean value (DAMV), peak mean (MPV), mean rectified electromyographic value (MRV), difference absolute standard deviation (DASDV), maximum fractal length (MFL), time-domain waveform length (TWL), power spectral density (PSD), phase locked value (PLV), standard deviation (STD), peak frequency (PEAK). Frequency (PKF), median frequency (MDF), and fourteen features.
[0207] Embodiment 16:
[0208] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Embodiments 11 to 15. Further, the calculation formula of the root mean square value RMS is as follows:
[0209]
[0210] Where i is the frequency point number, i = 1, 2, ..., N, N is the total number of frequency points. i is the electromyographic signal value at the i-th frequency point.
[0211] The calculation formula of the integrated electromyographic value IEMG is as follows:
[0212]
[0213] The calculation formula of the frequency domain waveform length FDWL is as follows:
[0214]
[0215] Where PSD(i) and PSD(i+1) are the power spectral densities of the electromyographic signal at the i-th and i+1-th frequency points, respectively, and △f is the frequency interval between adjacent frequency points.
[0216] The calculation formula of the difference absolute mean value DAMV is as follows:
[0217]
[0218] In the formula, x i+1 is the electromyographic signal value at the i+1th frequency point.
[0219] The calculation formula of the peak mean value MPV is as follows:
[0220]
[0221] Where j is the peak number, j = 1, 2, ..., m, and m is the total number of peaks. j is the jth peak in the electromyographic signal.
[0222] The calculation formula of the average rectified electromyographic value MRV is as follows:
[0223]
[0224] The calculation formula of the differential absolute standard deviation value DASDV is as follows:
[0225]
[0226] The calculation formula of the maximum fractal length MFL is as follows:
[0227] MFL=max(|x i+1 -x i |) (8)
[0228] The calculation formula of the time domain waveform length TWL is as follows:
[0229]
[0230] The power spectrum density is calculated by performing Fourier transform on the electromyographic signal in the frequency domain.
[0231] The phase locking value is obtained by calculation in the frequency domain.
[0232] The calculation formula of the standard deviation STD is as follows:
[0233]
[0234] In the formula, is the mean value of the electromyographic signal.
[0235] The peak frequency is determined by the peak in the frequency spectrum of the electromyographic signal.
[0236] The median frequency is determined by the median value in the frequency spectrum of the electromyographic signal.
[0237] Embodiment 17:
[0238] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 11 to 16. Furthermore, when the amputee imagines that his residual limb is still there and performs a variety of movements, the surface electromyographic signals of the four muscles of the forearm residual limb are extracted, and multiple features are extracted. The muscle activity analysis analyzes the features by visualization and comparison.
[0239] The muscle activity analysis method is to visualize the extracted features, display the activity intensity of different muscles in different actions and compare the activation levels of different muscles, so as to obtain the myoelectric level analysis of different actions.
[0240] The method for visualizing the extracted features includes but is not limited to comparing the RMS histograms of the surface electromyography of different muscles under a single action, comparing the surface electromyography IEMG histograms of muscles under different actions, comparing the radar charts of the surface electromyography FWL of different muscles under a single action, and comparing the radar charts of the surface electromyography RMS of different muscles under a single action.
[0241] Embodiment 18:
[0242] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 11 to 17. Furthermore, the muscle antagonism analysis is performed by calculating the antagonism coefficients of the radial side and the ulnar side, thereby obtaining an evaluation result of the antagonism between the agonist muscle and the antagonist muscle during wrist extension and wrist flexion.
[0243] The antagonism coefficient includes a wrist extension antagonism coefficient and a wrist flexion antagonism coefficient.
[0244] The muscle antagonism analysis characterizes the antagonism between the extensor and flexor muscles by calculating the antagonism coefficient. The antagonism reflects the patient's isolated muscle contraction ability. The stronger the antagonism, the better the ability to control the movement of the prosthesis.
[0245] The calculation formulas of the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are as follows:
[0246]
[0247]
[0248] Where, RMS屈肌 It is the sum of the root mean square values of the electromyographic signals of the radial flexor carpi and the ulnar flexor carpi in patients with forearm amputation. RMS 伸肌 It is the sum of the RMS values of the EMG signals of the radial extensor carpi and the ulnar extensor carpi in patients with forearm amputation.
[0249] Embodiment 19:
[0250] A muscle function assessment system for residual limbs of forearm amputees based on electromyographic signals, the main technical content of which is shown in any one of Embodiments 11 to 18. Further, the step of performing feature similarity analysis on the extracted features comprises:
[0251] 1) Standardize the extracted feature sequences and scale each standardized feature sequence using the L2 Euclidean norm. The scaled standardized feature sequence τ n As shown below:
[0252]
[0253] Where n is the sequence number of the electromyographic feature, ||T n || 2 represents the L2 Euclidean norm. mean(T n ) is the electromyographic characteristic sequence T n The mean of .
[0254] Among them, the nth electromyographic feature sequence T n As shown below:
[0255] T n =[t1, t2, ..., t g ] (14)
[0256] In the formula, t k is the electromyographic feature corresponding to the kth action, k = 1, 2, ..., g, where g is the total number of actions.
[0257] 2) All scaled and standardized feature sequences τ n Combined into a set of electromyographic signal feature sequences F corresponding to different muscle positions p , as shown below:
[0258] F p =[τ1,τ2,...,τ d ] (15)
[0259] Where d is the total number of myoelectric feature sequences and p represents the muscle position number of the forearm.
[0260] 3) The Pearson correlation coefficient was used to analyze the correlation between the characteristic sequences of myoelectric signals at different muscle positions of forearm amputees and healthy subjects, so as to describe the difference between the functional state of residual limb muscles of amputees and that of healthy subjects, as shown below:
[0261]
[0262] Where, P is the correlation coefficient between the EMG signal characteristics of the residual limb of the amputee and the EMG signal characteristics of the healthy subjects, P∈[-1,1]. p=1,2,...,q, q is the total number of muscle positions in the forearm. Represents the mean values of the characteristic sequences of electromyographic signals of different muscle positions in the forearm of patients with forearm amputation and healthy subjects. p Represents the characteristic sequence of electromyographic signals at the pth muscle position in the forearm of a healthy subject.
[0263] 4) The feature similarity analysis is combined with the clinical physician's evaluation results of the subjects and the feature similarity analysis results, and a set of muscle function evaluation indicators are obtained according to the distribution of the maximum correlation coefficient. The muscle function correlation coefficient of the residual limb of the amputee is mapped through the formula to obtain the evaluation result.
[0264] According to the similarity of the electromyographic signal characteristics of the residual limbs of amputees, the muscle function evaluation index is obtained as follows:
[0265]
[0266] Embodiment 20:
[0267] A muscle function assessment system for the residual limb of a forearm amputee based on electromyographic signals, the main technical content of which is shown in any one of Examples 11 to 19. Furthermore, the assessment report includes qualitative and quantitative information on muscle function obtained based on the results of data analysis and recommendations for prosthetic fitting and rehabilitation training formulated in combination with clinical judgment.
[0268] The evaluation report includes a report summary, experimental results, data analysis, and a summary.
[0269] The report summary includes the patient's name, age, gender, residual limb condition, amputation duration, etc.
[0270] The report summary includes the title and date of the report, as well as the patient's basic information, such as name, age, gender, residual limb condition, duration of amputation, etc.
[0271] The experimental results include the extracted characteristic values of multiple electromyographic signals, the antagonism coefficients of the extensor muscles and the flexor muscles, and the characteristic similarities.
[0272] The data analysis includes the myoelectric signals of different forearm muscles of forearm amputees completing each action, the antagonistic performance of agonist muscles and antagonist muscles, and muscle function evaluation indicators obtained by feature similarity analysis.
[0273] The data analysis includes analyzing the detailed information of the electromyographic signals and the degree of muscle activation of the patients in performing each task according to the experimental results, judging the antagonistic performance of the agonist and antagonist muscles according to the antagonistic coefficient, and judging the functional state of the residual limb muscles of the amputee according to the feature similarity.
[0274] The summary evaluates the myoelectric prosthesis control ability of the forearm amputee patients completing each action based on the analysis results of the myoelectric signals of different forearm muscles, the antagonistic performance of the agonist and antagonist muscles, and the muscle function evaluation indexes, and provides rehabilitation training suggestions.
[0275] If either the wrist extension antagonism coefficient or the wrist flexion antagonism coefficient is less than 0.5, and the muscle function evaluation index is weak or relatively weak, it is not suitable for fitting a myoelectric prosthesis. It is recommended to undergo residual limb rehabilitation training before re-evaluation.
[0276] If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is weak or relatively weak, it is recommended to conduct residual limb rehabilitation training before fitting the myoelectric prosthesis.
[0277] If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is strong or relatively strong, it is recommended to fit a myoelectric prosthesis.
[0278] The summary section includes evaluating whether the patient is suitable for fitting a two-degree-of-freedom myoelectric prosthesis based on the above data analysis results and in combination with the patient's lifestyle, rehabilitation goals and other factors, and providing rehabilitation training suggestions, including which tasks or movements to emphasize to improve muscle function and antagonist performance.
[0279] Embodiment 21:
[0280] See also Figures 1 to 5 A muscle function evaluation system for the residual limb of a forearm amputee based on electromyographic signals includes: a signal acquisition and transmission module, a signal processing module, and a signal analysis module.
[0281] The signal acquisition module collects the electromyographic signals of the residual limbs of forearm amputation patients under different movements, and transmits the data to the host computer software for subsequent signal processing and analysis.
[0282] The signal processing module performs preprocessing and feature extraction on the collected electromyographic data.
[0283] The signal analysis module performs muscle activity analysis, muscle antagonism analysis and feature similarity analysis according to the extracted features, and generates an evaluation report on the patient's residual limb muscle function.
[0284] The signal acquisition module uses electrodes made of Ag / Agcl (silver / silver chloride) to collect electromyographic signals of multiple muscles of the patient's residual limb, including but not limited to the flexor carpi ulnaris, flexor carpi radialis, extensor carpi ulnaris and extensor carpi radialis longus.
[0285] The signal acquisition module needs to collect the electromyographic signals of the amputee in seven movements, including but not limited to fist clenching, wrist extension, wrist flexion, wrist internal rotation, wrist external rotation, five-finger pinching, and palm extension.
[0286] The signal acquisition module sends the original electromyographic signal to the signal processing module via wireless transmission.
[0287] The signal processing module filters the original electromyographic signal through a 4th-order Butterworth bandpass filter with a passband range of 20 Hz-350 Hz to remove the high-frequency noise of the signal and the interference caused by poor contact between the electrode and the skin, and then uses a 4th-order Butterworth power frequency notch filter to filter out 50 Hz power frequency interference.
[0288] The signal processing module annotates different actions of the electromyographic signal and selects an analysis window with a window length of 200ms and a step length of 100ms to segment the multi-channel electromyographic signal in the activity segment, and obtains a series of analysis window data as a sample set for electromyographic feature extraction.
[0289] The signal processing module uses the above-screened sample set as the calculation input for feature extraction, and the various features include but are not limited to RMS (root mean square value), IEMG (integrated electromyographic value), FDWL (frequency domain waveform length), DAMV (difference absolute mean), MPV (peak mean), MRV (mean rectified electromyographic value), DASDV (differential absolute standard deviation), MFL (maximum fractal length), TWL (time domain waveform length), PSD power spectral density, PLV (phase locked value), STD (standard deviation), PKF (peak frequency), and MDF (median frequency).
[0290] The calculation method of RMS is shown in the following formula:
[0291]
[0292] Where N is the number of data points.
[0293] The calculation method of IEMG is as shown in the following formula:
[0294]
[0295] The calculation method of FDWL is as shown in the following formula:
[0296]
[0297] Where N is the number of frequency points in the spectrum, PSD(i) is the power spectral density of the signal at the i-th frequency point, and Δf is the frequency interval between adjacent frequency points.
[0298] The calculation method of DAMV is shown in the following formula:
[0299]
[0300] The calculation method of MPV is shown in the following formula:
[0301]
[0302] Where Pi is the peak value in the signal.
[0303] The calculation method of MRV is shown in the following formula:
[0304]
[0305] The calculation method of DASDV is shown in the following formula:
[0306]
[0307] The calculation method of MFL is shown in the following formula:
[0308] MFL=max(|x i+1 -x i ∣)
[0309] The calculation method of TWL is shown in the following formula:
[0310]
[0311] The PSD is calculated by performing Fourier transform on the EMG signal in the frequency domain and then calculating the power spectral density.
[0312] The PLV is calculated by calculating the phase lock value in the frequency domain.
[0313] The calculation method of STD is shown in the following formula:
[0314]
[0315] in is the mean of the signal.
[0316] The PKF is determined by the peaks in the frequency spectrum of the signal.
[0317] The MDF is determined by the median value in the frequency spectrum of the signal.
[0318] The signal analysis module uses the features for muscle activity analysis, muscle antagonism analysis and feature similarity analysis, and obtains a final evaluation report.
[0319] The muscle activity analysis analyzes the characteristics by visualization and comparison, including but not limited to the RMS histogram of the surface electromyography of different muscles of the patient under a single action, the surface electromyography IEMG histogram of muscles under different actions, the radar chart of the surface electromyography FWL of different muscles under a single action, the radar chart of the surface electromyography RMS of different muscles under a single action, etc., so as to intuitively observe the changes and differences in the electromyography characteristics of the residual limb muscles and the muscle activity patterns under different actions.
[0320] The muscle antagonism analysis characterizes the antagonism between the extensor and flexor muscles by calculating the antagonism coefficient, which reflects the patient's isolated muscle contraction ability. The stronger the antagonism, the better the ability to control the prosthesis movement. The antagonism coefficient calculation method is shown in the following formula:
[0321]
[0322]
[0323] The muscle antagonism analysis calculates the antagonism coefficients of the radial side and the ulnar side respectively, thereby obtaining an evaluation result of the antagonism between the agonist muscle and the antagonist muscle during wrist extension and wrist flexion movements.
[0324] The feature similarity analysis uses the Pearson correlation coefficient to analyze the correlation between the electromyographic signal feature sequences of different forearm muscles of the amputee and the healthy subject, and uses the correlation coefficient to evaluate the muscle function status of the residual limb of the amputee.
[0325] The feature similarity analysis first standardizes the feature sequence, and then uses the L2 Euclidean norm to scale each of the above standardized feature sequences in preparation for similarity analysis. The combination of a single feature sequence is shown in the following formula:
[0326] T n =[t1,t2,...,t l ]
[0327] Among them, n is the number of myoelectric features, and l is the number of collected actions.
[0328] The scaled and standardized feature sequence is shown in the following formula:
[0329]
[0330] Among them, ||T n ||2 represents the L2 norm of the input sequence.
[0331] The feature similarity analysis combines all the above-mentioned multiple standardized feature sequences to obtain a set of feature sequences corresponding to different muscle positions of the subjects, as shown in the formula:
[0332] F m =[τ1,τ2,...,τ n ]
[0333] Here, m represents the different muscle locations in the forearm.
[0334] The feature similarity analysis performs Pearson correlation coefficient analysis on the feature sequences of amputees and healthy subjects to describe the difference between the muscle function status of the residual limbs of amputees and the muscle function of healthy subjects. The Pearson correlation coefficient analysis process is shown in the following formula:
[0335]
[0336] The feature similarity analysis combines the evaluation results of the subjects by the clinical physicians and the analysis results of the feature similarity, and obtains a set of muscle function evaluation indicators according to the distribution of the maximum correlation coefficient. The muscle function correlation coefficient of the residual limb of the amputee is mapped through the formula to obtain the evaluation result. The muscle function evaluation index is shown in the following formula:
[0337]
[0338] The assessment report includes qualitative and quantitative information on muscle function obtained based on the results of data analysis and recommendations for prosthetic fitting and rehabilitation training formulated in combination with clinical judgment.
[0339] The contents of the evaluation report include: report summary, experimental results, data analysis, and summary.
[0340] The report summary includes the title and date of the report, as well as the patient's basic information, such as name, age, gender, residual limb condition, duration of amputation, etc.
[0341] The experimental results include the analyzed characteristic values of various electromyographic signals, the antagonism coefficients of the extensor and flexor muscles, and the characteristic similarities.
[0342] The data analysis includes analyzing the detailed information of the electromyographic signals and the degree of muscle activation of the patients in performing each task according to the experimental results, judging the antagonistic performance of the agonist and antagonist muscles according to the antagonistic coefficient, and judging the functional state of the residual limb muscles of the amputee according to the feature similarity.
[0343] The summary section includes evaluating whether the patient is suitable for fitting a two-degree-of-freedom myoelectric prosthesis based on the above data analysis results and in combination with the patient's lifestyle, rehabilitation goals and other factors, and providing rehabilitation training suggestions, including which tasks or movements to emphasize to improve muscle function and antagonist performance.
[0344] Embodiment 22:
[0345] See also Figures 1 to 5 , a muscle function evaluation system for the residual limb of a forearm amputee based on electromyographic signals, including a signal acquisition and transmission module, a signal processing module, and a signal analysis module.
[0346] The signal acquisition and transmission module acquires the electromyographic signals of the residual limbs of forearm amputation patients under different movements, and transmits the data to the host computer software for subsequent signal processing and analysis.
[0347] The signal processing module performs preprocessing and feature extraction on the collected electromyographic data.
[0348] The signal analysis module performs muscle activity analysis, muscle antagonism analysis and feature similarity analysis according to the extracted features, and generates an evaluation report on the patient's residual limb muscle function.
[0349] The signal acquisition module uses electrodes made of Ag / Agcl (silver / silver chloride) to collect electromyographic signals of multiple muscles of the patient's residual limb, including but not limited to the flexor carpi ulnaris, flexor carpi radialis, extensor carpi ulnaris and extensor carpi radialis longus.
[0350] The signal acquisition module needs to collect the electromyographic signals of the amputee in seven movements, including but not limited to fist clenching, wrist extension, wrist flexion, wrist internal rotation, wrist external rotation, five-finger pinching, and palm extension.
[0351] The signal acquisition module sends the original electromyographic signal to the signal processing module via wireless transmission.
[0352] The signal processing module filters the original electromyographic signal through a 4th-order Butterworth bandpass filter with a passband range of 20 Hz-350 Hz to remove the high-frequency noise of the signal and the interference caused by poor contact between the electrode and the skin, and then uses a 4th-order Butterworth power frequency notch filter to filter out 50 Hz power frequency interference.
[0353] The signal processing module annotates different actions of the electromyographic signal and selects an analysis window with a window length of 200ms and a step length of 100ms to further segment the multi-channel electromyographic signal in the activity segment, and obtains a series of analysis window data as a sample set for electromyographic feature extraction.
[0354] The signal processing module uses the above-screened sample set as the calculation input for feature extraction, and the various features include but are not limited to RMS (root mean square value), IEMG (integrated electromyographic value), FDWL (frequency domain waveform length), DAMV (difference absolute mean), MPV (peak mean), MRV (mean rectified electromyographic value), DASDV (differential absolute standard deviation), MFL (maximum fractal length), TWL (time domain waveform length), PSD power spectral density, PLV (phase locked value), STD (standard deviation), PKF (peak frequency), and MDF (median frequency).
[0355] The calculation method of RMS is shown in the following formula:
[0356]
[0357] Where N is the number of data points.
[0358] The calculation method of IEMG is as shown in the following formula:
[0359]
[0360] The calculation method of FDWL is as shown in the following formula:
[0361]
[0362] Where N is the number of frequency points in the spectrum, PSD(i) is the power spectral density of the signal at the i-th frequency point, and Δf is the frequency interval between adjacent frequency points.
[0363] The calculation method of DAMV is shown in the following formula:
[0364]
[0365] The calculation method of MPV is shown in the following formula:
[0366]
[0367] Where Pi is the peak value in the signal.
[0368] The calculation method of MRV is shown in the following formula:
[0369]
[0370] The calculation method of DASDV is shown in the following formula:
[0371]
[0372] The calculation method of MFL is shown in the following formula:
[0373] MFL=max(|x i+1 -x i ∣)
[0374] The calculation method of TWL is shown in the following formula:
[0375]
[0376] The PSD is calculated by performing Fourier transform on the EMG signal in the frequency domain and then calculating the power spectral density.
[0377] The PLV is calculated by calculating the phase lock value in the frequency domain.
[0378] The calculation method of STD is shown in the following formula:
[0379]
[0380] in is the mean of the signal.
[0381] The PKF is determined by the peaks in the frequency spectrum of the signal.
[0382] The MDF is determined by the median value in the frequency spectrum of the signal.
[0383] The signal analysis module performs analysis from the perspectives of muscle activity analysis, muscle antagonism analysis, and feature similarity analysis.
[0384] The muscle activity analysis analyzes the characteristics of the patient by visualization and comparison, including but not limited to the RMS histogram of the surface electromyography of different muscles of the patient under a single action, the surface electromyography IEMG histogram of muscles under different actions, the radar chart of the surface electromyography FWL of different muscles under a single action, the radar chart of the surface electromyography RMS of different muscles under a single action, etc., so as to intuitively observe the changes and differences in the electromyography characteristics of the residual limb muscles and the muscle activity patterns under different actions.
[0385] The muscle antagonism analysis characterizes the antagonism between the extensor and flexor muscles by calculating the antagonism coefficient, which reflects the patient's isolated muscle contraction ability. The stronger the antagonism, the better the ability to control the prosthesis movement. The antagonism coefficient calculation method is shown in the following formula:
[0386]
[0387]
[0388] The muscle antagonism analysis calculates the antagonism coefficients of the radial side and the ulnar side respectively, thereby obtaining an evaluation result of the antagonism between the agonist muscle and the antagonist muscle during wrist extension and wrist flexion movements.
[0389] The feature similarity analysis uses the Pearson correlation coefficient to analyze the correlation between the electromyographic signal feature sequences of different forearm muscles of the amputee and the healthy subject, and uses the correlation coefficient to evaluate the muscle function status of the residual limb of the amputee.
[0390] The feature similarity analysis first standardizes the feature sequence, and then uses the L2 Euclidean norm to scale each of the above standardized feature sequences in preparation for similarity analysis. The combination of a single feature sequence is shown in the following formula:
[0391] T n =[t1,t2,...,t l ]
[0392] Among them, n is the number of myoelectric features, and l is the number of collected actions.
[0393] The scaled and standardized feature sequence is shown in the following formula:
[0394]
[0395] Among them, ||T n || 2 Represents the L2 norm of the input sequence.
[0396] The feature similarity analysis combines all the above-mentioned multiple standardized feature sequences to obtain a set of feature sequences corresponding to different muscle positions of the subjects, as shown in the formula:
[0397] F m =[τ1,τ2,...,τ n ]
[0398] Here, m represents the different muscle locations in the forearm.
[0399] The feature similarity analysis performs Pearson correlation coefficient analysis on the feature sequences of amputees and healthy subjects to describe the difference between the muscle function status of the residual limbs of amputees and the muscle function of healthy subjects. The Pearson correlation coefficient analysis process is shown in the following formula.
[0400]
[0401] The feature similarity analysis combines the evaluation results of the subjects by the clinical physicians and the analysis results of the feature similarity, and obtains a set of muscle function evaluation indicators according to the distribution of the maximum correlation coefficient. The muscle function correlation coefficient of the residual limb of the amputee is mapped through the formula to obtain the evaluation result. The muscle function evaluation index is shown in the following formula:
[0402]
[0403] The assessment report includes qualitative and quantitative information on muscle function obtained based on the results of data analysis and recommendations for prosthetic fitting and rehabilitation training formulated in combination with clinical judgment.
[0404] The contents of the evaluation report include: report summary, experimental results, data analysis, and summary.
[0405] The report summary includes the title and date of the report, as well as the patient's basic information, such as name, age, gender, residual limb condition, duration of amputation, etc.
[0406] The experimental results include the analyzed characteristic values of various electromyographic signals, the antagonism coefficients of the extensor and flexor muscles, and the characteristic similarities.
[0407] The data analysis includes analyzing the detailed information of the electromyographic signals and the degree of muscle activation of the patients in performing each task according to the experimental results, judging the antagonistic performance of the agonist and antagonist muscles according to the antagonistic coefficient, and judging the functional state of the residual limb muscles of the amputee according to the feature similarity.
[0408] The summary section includes evaluating whether the patient is suitable for fitting a two-degree-of-freedom myoelectric prosthesis based on the above data analysis results and in combination with the patient's lifestyle, rehabilitation goals and other factors, and providing rehabilitation training suggestions, including which tasks or movements to emphasize to improve muscle function and antagonist performance.
Claims
1. A muscle function assessment system for residual limbs of forearm amputees based on electromyographic signals, characterized in that: include: Signal acquisition and transmission module, signal processing module, signal analysis module; The signal acquisition and transmission module is used to acquire the residual limb electromyographic signals of the forearm amputee in different actions, and transmit the residual limb electromyographic signals to the signal processing module; The signal processing module is used to preprocess and extract features of the collected residual limb electromyographic signals; The signal analysis module is used to perform muscle activity analysis, muscle antagonism analysis and feature similarity analysis on the extracted features, and generate an assessment report on the residual limb muscle function of forearm amputees; The step of performing feature similarity analysis on the extracted features comprises: 1) The extracted feature sequence is standardized, and each standardized feature sequence is scaled using the L2 Euclidean norm. The scaled standardized feature sequence τn is as follows: Where n is the sequence number of the electromyographic feature, ||T n ||2 represents the L2 Euclidean norm; mean(T n ) is the electromyographic characteristic sequence T n The mean of Among them, the nth electromyographic feature sequence T n As shown below: T n =[t1,t2,...,t g ] (2) Where, t k is the electromyographic feature corresponding to the k-th action, k = 1, 2, ..., g, g is the total number of actions; 2) All scaled and standardized feature sequences τn are combined into a set of EMG feature sequences F corresponding to different muscle positions p , as shown below: F p =[τ1,τ2,...,τ d ] (3) In the formula, d is the total number of myoelectric feature sequences; p represents the muscle position number of the forearm; 3) The Pearson correlation coefficient was used to analyze the correlation between the characteristic sequences of myoelectric signals at different muscle positions in the forearm of forearm amputees and healthy subjects, as shown below: Where, P is the correlation coefficient between the EMG signal characteristics of the residual limb of the amputee and the EMG signal characteristics of the healthy subjects, P∈[-1,1]; p=1,2,…,q, q is the total number of muscle positions in the forearm; Represents the mean values of the characteristic sequences of electromyographic signals of different muscle positions in the forearm of patients with forearm amputation and healthy subjects; y p The characteristic sequence of myoelectric signals representing the pth muscle position in the forearm of a healthy subject; 4) According to the similarity of the electromyographic signal characteristics of the residual limbs of amputees, the muscle function evaluation index is obtained as follows:
2. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The signal acquisition and transmission module acquires electromyographic signals of the residual limb from the following parts, including but not limited to the flexor carpi ulnaris, flexor carpi radialis, extensor carpi ulnaris, and extensor carpi radialis longus.
3. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The muscle function assessment system also includes a gesture guidance module; The gesture guidance module guides the patient to use the residual limb to complete the gesture action in the form of images and / or text instructions through the display, and the playback time of each gesture action is t 0c ; c represents the cth gesture action; the interval between each two gesture actions is t 0u ; The gestures include but are not limited to resting gestures, single-degree-of-freedom motions, and multi-degree-of-freedom motions; The single degree of freedom movement includes wrist extension, wrist flexion, wrist internal rotation, and wrist external rotation; The multi-degree-of-freedom movements include fist clenching, five-finger pinching, and palm extension; When the gesture action is a resting gesture action or a single-degree-of-freedom motion, the residual limb electromyographic signal is a single-degree-of-freedom motion electromyographic signal; When the gesture action is a multi-degree-of-freedom movement, the residual limb electromyographic signal is a multi-degree-of-freedom movement electromyographic signal, including electromyographic signals from at least two flexor muscles and / or extensor muscles of the residual limb.
4. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The preprocessing includes filtering and data segmentation; The data segmentation is used to select a window for feature extraction.
5. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The features include but are not limited to root mean square value, integrated EMG value, frequency domain waveform length, absolute mean value of difference, peak mean value, average rectified EMG value, absolute standard deviation value of difference, maximum fractal length, time domain waveform length, power spectral density, phase lock value, standard deviation, peak frequency, and median frequency.
6. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 5, characterized in that: The calculation formula of the root mean square value RMS is as follows: Where i is the frequency point number, i = 1, 2, ..., N, N is the total number of frequency points; x i is the electromyographic signal value at the i-th frequency point; The calculation formula of the integrated electromyographic value IEMG is as follows: The calculation formula of the frequency domain waveform length FDWL is as follows: In the formula, PSD(i) and PSD(i+1) are the power spectral densities of the electromyographic signal at the i-th and i+1-th frequency points, respectively, and Δf is the frequency interval between adjacent frequency points; The calculation formula of the difference absolute mean value DAMV is as follows: In the formula, x i+1 is the electromyographic signal value at the i+1th frequency point; The calculation formula of the peak mean value MPV is as follows: Where j is the peak number, j = 1, 2, ..., m, m is the total number of peaks; P j is the jth peak value in the electromyographic signal; The calculation formula of the average rectified electromyographic value MRV is as follows: The calculation formula of the differential absolute standard deviation value DASDV is as follows: The calculation formula of the maximum fractal length MFL is as follows: MFL=max(∣x i+1 -x i ∣) (13) The calculation formula of the time domain waveform length TWL is as follows: The power spectrum density is calculated by performing Fourier transform on the electromyographic signal in the frequency domain; The phase locking value is obtained by calculation in the frequency domain; The calculation formula of the standard deviation STD is as follows: In the formula, is the mean value of the electromyographic signal; The peak frequency is determined by the peak value in the spectrum of the electromyographic signal; The median frequency is determined by the median value in the frequency spectrum of the electromyographic signal.
7. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The muscle activity analysis method is to visualize the extracted features, display the activity intensity of different muscles in different actions and compare the activation levels of different muscles, so as to obtain the electromyographic level analysis of different actions; The method for visualizing the extracted features includes but is not limited to comparing the RMS histograms of the surface electromyography of different muscles under a single action, comparing the surface electromyography IEMG histograms of muscles under different actions, comparing the radar charts of the surface electromyography FWL of different muscles under a single action, and comparing the radar charts of the surface electromyography RMS of different muscles under a single action.
8. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The muscle antagonism analysis is to calculate the antagonism coefficients of the radial side and the ulnar side, thereby obtaining an evaluation result of the antagonism between the agonist muscle and the antagonist muscle during wrist extension and wrist flexion movements; The antagonism coefficient includes a wrist extension antagonism coefficient and a wrist flexion antagonism coefficient; The calculation formulas of the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are as follows: Where, RMS 屈肌 RMS is the sum of the root mean square values of the electromyographic signals of the radial flexor carpi and the ulnar flexor carpi in patients with forearm amputation. 伸肌 It is the sum of the RMS values of the EMG signals of the radial extensor carpi and the ulnar extensor carpi in patients with forearm amputation.
9. The muscle function evaluation system for residual limbs of forearm amputees based on electromyographic signals according to claim 1, characterized in that: The evaluation report includes a report summary, experimental results, data analysis, and summary; The report summary includes the patient's name, age, sex, residual limb condition, and duration of amputation; The experimental results include the extracted characteristic values of multiple electromyographic signals, the antagonism coefficients of the extensor and flexor muscles, and the characteristic similarity; The data analysis includes the electromyographic signals of different forearm muscles of forearm amputees completing each action, the antagonistic performance of agonist and antagonist muscles, and the muscle function evaluation index obtained by feature similarity analysis; The summary evaluates the myoelectric prosthesis control ability of the forearm amputee according to the analysis results of the myoelectric signals of different forearm muscles, the antagonistic performance of the agonist and antagonist muscles, and the muscle function evaluation indexes when the forearm amputee completes each action, and gives rehabilitation training suggestions; If either the wrist extension antagonism coefficient or the wrist flexion antagonism coefficient is less than 0.5, and the muscle function evaluation index is weak or relatively weak, it is not suitable to fit the myoelectric prosthesis. It is recommended to undergo residual limb rehabilitation training and then re-evaluate. If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is weak or relatively weak, it is recommended to first conduct residual limb rehabilitation training and then fit the myoelectric prosthesis; If the wrist extension antagonism coefficient and the wrist flexion antagonism coefficient are both greater than 0.5, and the muscle function evaluation index is strong or relatively strong, it is recommended to fit a myoelectric prosthesis.
