Individualized functional electrical stimulation system based on muscle cooperative compensation model

Abstract

The invention discloses a personalized functional electrical stimulation system based on a muscle cooperative compensation model. Comprising a data acquisition and preprocessing module, a muscle activation coefficient compensation module and an electrical stimulation module. The method comprises the following steps of: analyzing an electromyographic signal and a motion posture angle signal of a limb on an uninjured side, extracting a muscle coordination structure on the uninjured side, establishing an injured side muscle activation compensation model under the guidance of the uninjured side coordination structure, and customizing a personalized electrical stimulation strategy by utilizing the compensation model and the muscle coordination structure on the uninjured side. According to the invention, the cooperative structure of the healthy side of the patient is utilized to guide the affected side to move, and electrical stimulation compensation with corresponding intensity can be performed according to the difference of the muscle activation degrees of the healthy side and the affected side in different motion states, so that the damaged muscle group can be recovered in a more targeted manner, and the neuroplasticity of the central system can be improved; and the working efficiency of the electrical stimulator is also improved by extracting the cooperative structure.

Description

Technical Field

[0001] The present invention belongs to the field of functional electrical stimulation medical rehabilitation, and specifically relates to a personalized functional electrical stimulation system based on a muscle synergy compensation model. Background Art

[0002] The muscle synergy theory states that movement is initiated by the brain controlling a group of synergistic modules with specific functions located in the motor cortex and spinal cord, and each muscle synergy simultaneously activates a group of muscles with different weights. Neurological diseases such as stroke are often accompanied by changes in the number, composition, activation coefficient and recruitment method of muscle synergies. Muscle synergy can be used for functional electrical stimulation for rehabilitation treatment.

[0003] Functional electrical stimulation is a commonly used intervention method for motor rehabilitation of post-stroke patients. It can promote motor recovery and improve the neuroplasticity of the central nervous system by applying low-frequency pulse current to the corresponding muscles or muscle groups. Most functional electrical stimulation strategies usually fix the intensity of the stimulation parameters or manually adjust the stimulation parameters based on experience. However, neither the fixed stimulation parameters nor the manual adjustment methods take into account the differences between individuals and the different degrees of motor impairment after stroke, which makes the electrical stimulation therapy fail to achieve the desired effect and even cause certain damage to the muscles. Therefore, in order to carry out electrical stimulation therapy more effectively, it is necessary to formulate an individualized electrical stimulation strategy based on the patient's own characteristics.

[0004] The prior Chinese patent "An electrical stimulation method and system based on healthy lower limb muscle synergy inspiration" proposed to construct a motion guidance model by extracting the synergy relationship of the healthy limb muscles of an individual, and set electrical stimulation parameters according to the motion guidance model. However, it only takes into account the differences in muscle synergy relationships between different individuals, so the accuracy of its electrical stimulation scheme is still insufficient. Summary of the invention

[0005] In view of the shortcomings of existing electrical stimulation strategy settings, the present invention proposes a personalized functional electrical stimulation system based on a muscle synergy compensation model. Taking into account the actual motor injury of the affected limb after stroke or the recovery of the muscles of the affected limb during rehabilitation training, more effective electrical stimulation compensation is performed on the muscles of the affected limb by comparing the difference in activation levels between the affected and healthy limbs.

[0006] The technical solutions specifically adopted in the present invention are as follows: A personalized functional electrical stimulation system based on a muscle synergistic compensation model, comprising: The data acquisition and preprocessing module is responsible for collecting the electromyographic signals and motion posture angle signals of the healthy and affected limbs; preprocessing the electromyographic signals, then resampling the electromyographic signals and mapping them to the motion posture angle space, and finally segmenting the resampled electromyographic signals into motion cycles to obtain the electromyographic signals that change with the motion posture angle in each motion cycle; The muscle activation coefficient compensation module is responsible for extracting muscle synergy from the electromyographic signal of the healthy limb after period segmentation, obtaining the muscle synergy structure matrix and muscle activation coefficient matrix of the healthy limb, and using them as the guidance model; according to the guidance model, the muscle activation coefficient matrix of the affected limb is calculated, and the muscle activation compensation model of the affected limb is established; The electrical stimulation module is responsible for setting the multi-channel electrical stimulation parameters of the electrical stimulator and performing electrical stimulation according to the muscle activation compensation model of the affected limb and the muscle synergy structure matrix of the healthy limb.

[0007] Based on the similarity and symmetry of the muscle coordination structure and movement posture of the left and right limbs of healthy people, the present invention proposes the idea of ​​using the coordination structure of the healthy limb to guide the movement rehabilitation of the affected limb, providing users with a rehabilitation strategy that better suits their own coordination conditions.

[0008] The present invention also takes into account the differences in the degree of stroke and motor injury in patients, extracts the activation coefficient of the affected side under the guidance of the healthy side collaborative structure, and establishes a muscle activation compensation model based on the difference in activation coefficients between the healthy and affected sides, providing a basis for implementing the electrical stimulation compensation strategy of "how much is lacking, how much is supplemented".

[0009] The present invention simplifies the setting of stimulation parameters, extracts the collaborative structure through matrix decomposition, and obtains the weight ratio of each muscle under the collaborative structure. There is no need to set the parameter value of each stimulation channel separately. Instead, according to the muscle participation ratio and the required compensation stimulation angle, the dimensionality reduction control of the functional electrical stimulator is achieved and the influence of the uncertainty of the exercise time required for different patients in rehabilitation training is avoided, thereby improving work efficiency.

[0010] In summary, the core of the present invention is to utilize the natural symmetry of the muscle synergy structure and movement posture of the left and right limbs of a healthy human body, and provide an innovative sports rehabilitation guiding ideology. This idea provides personalized guidance for the rehabilitation exercise of the affected limb based on the normal synergy mode of the healthy limb, so as to better fit the patient's own coordination characteristics and achieve more accurate and effective rehabilitation training. In view of the difference in stroke degree and sports injury of different patients, the present invention extracts the activation coefficient of the affected limb through the guidance of the healthy synergy structure, and constructs a muscle activation compensation model based on the difference in the activation coefficient of the healthy and affected sides. The present invention not only extracts the muscle synergy structure of the healthy side, but also establishes an affected side muscle activation compensation model based on the healthy synergy structure through in-depth analysis of the electromyographic signals and movement posture angle signals of the healthy and affected limbs of the patient. Utilizing this model and the healthy synergy structure, the present invention can customize personalized electrical stimulation strategies. The present invention can not only promote the recovery of damaged muscle groups in a targeted manner through the guidance of the healthy synergy structure and personalized electrical stimulation compensation, but also improve the neuroplasticity of the central nervous system, thereby accelerating the patient's rehabilitation process. At the same time, by improving the working efficiency of the electrical stimulator, the present invention also brings important technological progress and innovative value to the field of rehabilitation medicine. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is a flow chart of the present invention; Figure 2 Flowchart for establishing the muscle activation coefficient compensation model; Figure 3 The muscle synergy structure matrix and activation coefficient matrix of the healthy limb obtained after NMF decomposition in this embodiment, where (a) is the muscle weight distribution in the first muscle synergy mode, (b) is the activation curve with angle change in the first muscle synergy mode, (c) is the muscle weight distribution in the second muscle synergy mode, and (d) is the activation curve with angle change in the second muscle synergy mode; Figure 4 : are muscle activation compensation curves of the affected limb in this embodiment, wherein (a) is a muscle activation compensation curve that changes with angle under the guidance of the first muscle synergy pattern of the healthy limb, and (b) is a muscle activation compensation curve that changes with angle under the guidance of the second muscle synergy pattern of the healthy limb; Figure 5 It is the electrical stimulation amplitude intensity of each muscle of the affected limb in this embodiment, including the rectus femoris VR, lateral thigh muscle VL, biceps thigh muscle BM, semitendinosus muscle ST, tibialis anterior muscle AT, peroneus longus PL, medial gastrocnemius muscle GM, and soleus muscle SM. DETAILED DESCRIPTION

[0012] The method of the present invention is further described in detail below in conjunction with the accompanying drawings.

[0013] The present invention provides a personalized functional electrical stimulation system based on a muscle synergy compensation model, such as Figure 1 include: The data acquisition and preprocessing module is responsible for collecting the electromyographic signals and motion posture angle signals of the healthy and affected limbs; preprocessing the electromyographic signals, then resampling the electromyographic signals and mapping them to the motion posture angle space, and finally segmenting the resampled electromyographic signals into motion cycles to obtain the electromyographic signals that change with the motion posture angle in each motion cycle; The muscle activation coefficient compensation module is responsible for extracting muscle synergy from the electromyographic signal of the healthy limb after period segmentation, obtaining the muscle synergy structure matrix and muscle activation coefficient matrix of the healthy limb, and using them as the guidance model; according to the guidance model, the muscle activation coefficient matrix of the affected limb is calculated, and the muscle activation compensation model of the affected limb is established; The electrical stimulation module is responsible for setting the multi-channel electrical stimulation parameters of the electrical stimulator and performing electrical stimulation according to the muscle activation compensation model of the affected limb and the muscle synergy structure matrix of the healthy limb.

[0014] Specifically, the EMG signals are preprocessed in the data acquisition and preprocessing module as follows: the surface EMG signals collected from the healthy limb and the affected limb are subjected to 50 Hz power frequency filtering, 50-450 Hz band-pass filtering, and 10 Hz low-pass filtering to obtain their respective EMG signal envelopes.

[0015] Specifically, in the data acquisition and preprocessing module, the preprocessed electromyographic signal is resampled and mapped to the motion posture angle space, and finally the resampled electromyographic signal is segmented into motion cycles to obtain the electromyographic signal that changes with the motion posture angle in each motion cycle; specifically: First, the motion posture angle signals of the healthy limb and the affected limb are collected, and then the motion posture angle signals of the same limb are aligned with the preprocessed electromyographic signal envelope; that is, considering the uneven speed of angle change in the motion state, the electromyographic signals within a certain range of angle change are resampled every time the motion posture angle changes within a certain range; based on 100 points every 10 degrees, the electromyographic signals that change with time are mapped accordingly to the motion posture angle range.

[0016] Then, the electromyographic signal is segmented into motion cycles so that the length of the electromyographic signal in each motion cycle is consistent.

[0017] Specifically, the muscle synergy extraction is performed on the electromyographic signal of the healthy limb after periodic segmentation in the muscle activation coefficient compensation module to obtain the muscle synergy structure matrix and muscle activation coefficient matrix of the healthy limb, and use them as the guiding model; specifically: like Figure 2As shown in the figure, the muscle synergy information is obtained by using the non-negative matrix factorization (NMF) algorithm for the electromyographic data of the healthy limb after segmentation according to the movement cycle. The decomposition formula is expressed as: Wherein, V is the electromyographic data of the healthy limb after the movement cycle is segmented, m represents the number of channels of the electromyographic signal, and n represents the number of sampling points in the movement cycle; W is the muscle synergy structure matrix, each column represents the weight distribution of a group of synergies in each muscle, H is the muscle activation coefficient matrix, each row represents a group of activation coefficients that change with the angle, and r represents the number of muscle synergies.

[0018] The steps of the NMF algorithm are: Initialize W and H randomly and all elements in the matrix need to be non-negative.

[0019] Minimize the objective function: Use the gradient descent method to iterate W and H: Repeat the iterative update steps until the convergence condition is met, such as reaching the maximum number of iterations or the objective function is less than a certain threshold. Determine the final muscle synergy structure matrix W and muscle activation coefficient matrix H.

[0020] Before executing the NMF algorithm, it is necessary to determine the value of r, that is, the number of muscle synergies. This value can be determined by calculating the value of VAF, specifically: Determine the number of synergies r, which gradually increases from 1 to the number of myoelectric signal channels m, and then calculate the reconstruction accuracy VAF of the decomposed matrices W and H to the original matrix V: The closer the VAF is to 1, the more accurate the reconstructed matrix is.

[0021] The VAF under each motion cycle is averaged. When the VAF mean is greater than the set threshold θ, such as 0.95, and the number of synergies r is increased by one so that the VAF mean increases to less than a threshold β, such as 0.05, r is determined to be the final synergy number.

[0022] After NMF decomposition, Figure 3 The muscle synergy structure matrix and activation coefficient matrix of the healthy limb were obtained, where the synergy number r was 2. (rectus femoris VR, lateral thigh muscle VL, biceps thigh muscle BM, semitendinosus muscle ST, tibialis anterior muscle AT, peroneus longus PL, medial gastrocnemius muscle GM, soleus muscle SM).

[0023] Specifically, in the muscle activation coefficient compensation module, the muscle activation coefficient matrix of the affected limb is calculated according to the guidance model, and the muscle activation compensation model of the affected limb is established; Figure 2 Specifically: (1) Based on the muscle synergy structure matrix obtained from the healthy limb, the muscle activation coefficient matrix of the affected limb is extracted from the electromyographic signal of the affected limb: in, is the electromyographic data after the segmentation of the movement cycle of the affected limb, m represents the number of channels of the electromyographic signal, and n represents the number of sampling points of the movement cycle; W is the muscle synergy structure matrix of the healthy side, r represents the number of muscle synergies, is the muscle activation coefficient matrix of the affected side, and each row represents a set of activation coefficients that vary with angle.

[0024] The muscle activation coefficient matrix of the affected side is calculated by the least square method; The least squares method minimizes the objective function: Construct the equation to ensure that A is reversible: Then the muscle activation coefficient matrix of the affected side is : (2) Subtract the muscle activation coefficient matrices of the healthy limb and the affected limb to obtain the compensation matrix D. Finally, normalize the compensation matrix D to obtain the muscle activation compensation model of the affected limb: Specifically, in the electrical stimulation module, the multi-channel electrical stimulation parameters of the electrical stimulator are set and the electrical stimulation is performed according to the muscle activation compensation model of the affected limb and the muscle synergy structure matrix of the healthy limb. The stimulation parameters of the electrical stimulator are set according to the muscle activation compensation model of the affected limb, and a personalized functional electrical stimulation strategy is formulated; specifically: (1) According to the muscle activation compensation model of the affected limb, the affected limb muscle activation compensation curve is obtained, such as Figure 4 ;Threshold the muscle activation compensation curve on the affected side, according to the set activation threshold The data value is converted to 0 or 1 to obtain the converted matrix Q. Specifically, if the data value is greater than or equal to the threshold, it is set to 1; if the data value is less than the threshold, it is set to 0. The motion posture angle range corresponding to the value of 1 in the converted matrix Q is the electrical stimulation range of the electrical stimulator.

[0025] (2) The healthy side muscle synergy structure matrix W obtained in step 2 is multiplied by the transformed matrix Q and normalized to obtain the matrix F, which is the compensation weight and electrical stimulation range of each muscle on the affected limb.

[0026] (3) Set the frequency and pulse width of the electrical stimulator, and change the amplitude of the electrical stimulation. Figure 5 Electrical stimulation compensation, see the following formula: in, It indicates the maximum stimulation intensity that the patient can accept. represents the electrical stimulation intensity to be applied to the i-th muscle of the affected limb at the movement posture angle t, represents the value with the largest weight in the matrix F, represents the value of the i-th muscle of the affected limb at the movement posture angle t, A value of 0 means that no compensation is required at angle t, that is, no electrical stimulation is required.

[0027] According to the set stimulation parameters (frequency value, pulse width value, electrical stimulation amplitude intensity), When the angle of the movement posture is not 0, the electric stimulator is started to apply electric stimulation. When it is 0, the electrical stimulator stops working.

[0028] An embodiment of the present invention provides an electronic device. Specifically, the electronic device includes a memory and a processor. The memory stores executable code. When the processor executes the executable code, the method described in any one of the implementation modes is implemented.

[0029] The memory may include high-speed random access memory (RAM), and may also include non-volatile memory (Non-volatile Memory), such as at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0030] The bus may be an ISA bus, a PCI bus or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc.

[0031] Among them, the memory is used to store the program, and the processor executes the program after receiving the execution instruction. The method executed by the device for flow process definition disclosed in any embodiment of the above-mentioned embodiment of the present invention can be applied to the processor or implemented by the processor.

[0032] The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present invention can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in a decoding processor. The software module may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0033] The computer program product of the readable storage medium provided in the embodiment of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the methods described in the previous method embodiments. The specific implementation can be referred to the previous method embodiments, which will not be repeated here.

[0034] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0035] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A personalized functional electrical stimulation system based on a muscle synergy compensation model, characterized in that include: The data acquisition and preprocessing module is responsible for collecting the electromyographic signals and movement posture angle signals of the healthy and affected limbs; Preprocess the electromyographic signal, then resample the electromyographic signal and map it to the motion posture angle space, and finally segment the resampled electromyographic signal into motion cycles to obtain the electromyographic signal that changes with the motion posture angle in each motion cycle; The muscle activation coefficient compensation module is responsible for extracting muscle synergy from the electromyographic signals of the healthy limb after period segmentation, obtaining the muscle synergy structure matrix and muscle activation coefficient matrix of the healthy limb, and using them as a guiding model; According to the guidance model, the muscle activation coefficient matrix of the affected limb is calculated, and the muscle activation compensation model of the affected limb is established; The electrical stimulation module is responsible for setting the multi-channel electrical stimulation parameters of the electrical stimulator and performing electrical stimulation according to the muscle activation compensation model of the affected limb and the muscle synergy structure matrix of the healthy limb.

2. The system according to claim 1, characterized in that: The preprocessing of electromyographic signals includes power frequency removal filtering, bandpass filtering and low-pass filtering.

3. The system according to claim 1, characterized in that: The resampling and mapping of the electromyographic signal to the motion posture angle space is to align the motion posture angle signal of the same limb with the electromyographic signal envelope obtained after preprocessing the electromyographic signal.

4. The system according to claim 1, characterized in that: The guiding model building process in the muscle activation coefficient compensation module is specifically: The muscle synergy information is obtained by using the non-negative matrix decomposition algorithm on the electromyographic data of the healthy limb segmented according to the movement cycle; The muscle synergy number r in the non-negative matrix decomposition algorithm is determined by calculating the reconstruction accuracy of the healthy limb electromyographic data V after periodic segmentation by calculating the muscle synergy structure matrix W and the muscle activation coefficient matrix H after decomposition, specifically: Determine the number of synergies r, which gradually increases from 1 to the number of myoelectric signal channels m, and then calculate the reconstruction accuracy VAF of the decomposed muscle synergy structure matrix W and the muscle activation coefficient matrix H for the healthy limb electromyographic data V after periodic segmentation; The average VAF of each period under the current collaborative number r is calculated. When the VAF average is greater than the set threshold θ, θ>0, and the increase in the VAF average when the collaborative number r is increased by one is less than the threshold β, β>0, the current collaborative number r is determined to be the final collaborative number.

5. The system according to claim 4, characterized in that: The calculation of the reconstruction accuracy VAF in the muscle activation coefficient compensation module that guides the model building process is as follows: Formula (1).

6. The system according to claim 1, characterized in that: The specific process of constructing the muscle activation compensation model of the affected limb is: Based on the muscle synergy structure matrix obtained from the healthy limb, the muscle activation coefficient matrix of the affected limb is extracted from the electromyographic signal of the affected limb: Formula (2) in, is the electromyographic data after the segmentation of the movement cycle of the affected limb, m represents the number of channels of the electromyographic signal, and n represents the number of sampling points of the movement cycle; W is the muscle synergy structure matrix of the healthy limb, r represents the number of muscle synergies, is the muscle activation coefficient matrix of the affected limb, each row represents a set of activation coefficients that change synergistically with the angle of the movement posture; The muscle activation coefficient matrices of the healthy limb and the affected limb are subtracted to obtain the compensation matrix D; finally, the compensation matrix D is normalized to obtain the muscle activation compensation model of the affected limb.

7. The system according to claim 6, characterized in that: During the construction of the muscle activation compensation model of the affected limb, the muscle activation coefficient matrix of the affected limb is calculated by the least squares method.

8. The system according to claim 1, characterized in that: The specific implementation process of the electrical stimulation module is: According to the muscle activation compensation model of the affected limb, the affected-side muscle activation compensation curve is obtained; the affected-side muscle activation compensation curve is thresholded, and the data value is converted to 0 or 1 to obtain the converted matrix Q; the movement posture angle range corresponding to the value of 1 in the converted matrix Q is the electrical stimulation range of the electrical stimulator; The matrix F is obtained by multiplying the muscle synergy structure matrix W of the healthy limb and the transformed matrix Q, and then normalizing it, which is the compensation weight of each muscle on the affected limb. Set the frequency and pulse width of the electrical stimulator, and perform electrical stimulation compensation by changing the electrical stimulation amplitude intensity, as shown in the following formula: Formula (3) in, It indicates the maximum stimulation intensity that the patient can accept. represents the electrical stimulation amplitude intensity to be applied to the i-th muscle of the affected limb at the movement posture angle t, represents the value with the largest compensation weight in the matrix F, represents the value of the i-th muscle of the affected limb at the movement posture angle t, If it is 0, it means that no compensation is required at the motion posture angle t, that is, no electrical stimulation is required; According to the frequency value, pulse width value, and amplitude intensity of the electrical stimulation instrument ,exist When the angle of the movement posture is not 0, the electric stimulator is started to apply electric stimulation. When it is 0, the electrical stimulator stops working.

9. An electronic device, characterized in that: The system comprises a processor and a memory, wherein the memory stores machine executable instructions that can be executed by the processor, and the processor executes the machine executable instructions to implement the system according to any one of claims 1 to 8.

10. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are called and executed by a processor, the machine-executable instructions prompt the processor to implement the system described in any one of claims 1-8.

Images