An electromyographic signal processing method, system and device for stroke patient rehabilitation training

Abstract

The application discloses a kind of electromyographic signal processing method, system and equipment for stroke patient rehabilitation training, belong to rehabilitation medical electronics technical field.Method includes: through three acquisition point serpentine grid flexible FPC electrode to obtain differential signal and reference signal, RF interference is filtered out by RFI filter;Signal is processed using three operational amplifier instrument amplifiers: first stage instrument amplifier suppresses common-mode interference and realizes primary filtering, second stage two-order active filter constitutes three-pole bandpass characteristic, and output signal amplitude is amplified to 2 / 3 of ADC range accurately;Direct current bias voltage generated by reference signal is superimposed with amplified signal, and bipolar signal is stabilized to the input range of adaptive ADC is converted by 24-bit ADC;It is transmitted to terminal by self-defined bluetooth protocol, and real-time generation electromyographic power spectral density analysis and rehabilitation quantization report.The technical scheme of the present application can effectively limit the signal band in the main energy range of electromyographic signal, filter out high-frequency noise and motion artifacts in addition to muscle signal.

Description

Technical Field

[0001] This invention relates to the field of electronic processing technology in rehabilitation medicine, and in particular to a method, system and device for electromyography signal processing for rehabilitation training of stroke patients. Background Technology

[0002] Wearable health monitoring devices are increasingly used in elderly patients' physical monitoring, rehabilitation monitoring, and sports training. In rehabilitation training, quantitative analysis often requires significant medical resources and human support. Because the degree of rehabilitation is difficult to quantify, it often relies on the physician's experience, leading to a continuous increase in the demand for rehabilitation status monitoring devices. However, due to the varying locations of affected areas and generally high muscle tension among patients, traditional steel electrodes cannot provide a close fit.

[0003] With societal development, the incidence of brain diseases such as stroke is constantly rising. These diseases are often accompanied by symptoms such as upper body motor difficulties, making postoperative rehabilitation particularly challenging. Rehabilitation training is an integral part of the patient's long recovery process. However, current rehabilitation methods suffer from problems such as the inability to quantify and analyze data, and the inability to determine the appropriate training volume.

[0004] Electromyography (EMG) signals can visually display movement patterns during patient rehabilitation and have significant reference value. Through EMG, rehabilitation levels can be assessed, recovery stages can be defined, training volume calculated, and training schedules scheduled. Human muscle EMG signal acquisition devices have become increasingly important in modern medicine due to their reliable diagnostic capabilities, ease of operation, and harmlessness.

[0005] In current clinical applications, silver chloride wet electrodes are commonly used as the most distal acquisition device for human electromyography (EMG) signals. To reduce poor electrode contact caused by movement deformation and to ensure stability during acquisition, it is usually necessary to wipe the electrode with alcohol, apply conductive adhesive, and then attach the silver chloride electrode before acquisition. This process not only causes inconvenience to patients and doctors, but the conductive adhesive is also prone to drying and curing, leading to excessive noise during long-term signal acquisition.

[0006] With advancements in technology, the fabrication of dry electrodes has become possible, enabling long-term monitoring of the human body. Dry electrodes typically consist of several parts: a conductive portion made of a low-resistivity metal, a substrate usually made of a flexible material to ensure skin adhesion, and a wire portion typically made of copper. However, traditional flexible dry electrodes often suffer from high cost, complex manufacturing processes, and non-standard interfaces. To address these issues, this invention proposes a dry electrode fabrication method based on a combination of FPC flexible connecting wires and screen printing technology. This method, while maintaining a certain degree of flexibility, significantly simplifies the electrode fabrication process, improves standardization, enhances the stability of the finished product, and offers good customizability.

[0007] Generally speaking, there are five types of interference noise on the surface of the human body: 1. Interference caused by electrocardiogram signals; 2. Motion interference caused by skin deformation; 3. Motion interference caused by wires; 4. RF interference caused by external radio frequency signals; 5. Motion artifacts caused by friction between metal and skin.

[0008] In summary, how to eliminate the impact of these interferences on the acquired signals is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0009] The technical solution of this invention to solve the above-mentioned technical problems is to provide a method for electromyographic signal processing for rehabilitation training of stroke patients, comprising the following steps:

[0010] S1. Signal Acquisition and RF Filtering:

[0011] Differential electrode signals (IN+, IN) and reference electrode signals (REF) are acquired using a three-point flexible FPC electrode.

[0012] The three signals are input into the RFI radio frequency filter to filter out high-frequency radio frequency interference;

[0013] S2. Differential Amplification and Three-Stage Filtering:

[0014] The filtered differential signals (IN+, IN) are input into a three-op-amp instrumentation amplifier:

[0015] First stage: The instrumentation amplifier structure is adopted to amplify the differential signal and suppress common-mode interference, while realizing the first stage of low-pass filtering;

[0016] Second stage: High-order filtering is performed through a second-order active filter, which together with the first stage forms a three-pole bandpass characteristic;

[0017] Amplify the output signal amplitude to 2 / 3 of the maximum value of the ADC range;

[0018] S3. Signal boosting and ADC conversion:

[0019] Process the reference electrode signal (REF) to generate a DC bias voltage;

[0020] The DC bias voltage is added to the amplified differential signal to stabilize the bipolar electromyographic signal to the input range of the ADC.

[0021] Analog signals are converted into digital signals using a 24-bit ADC signal converter;

[0022] S4. Data Calibration and Transmission:

[0023] Perform a calibration algorithm on the ADC output data:

[0024] 1) Set the encoded value representing the negative full scale to zero;

[0025] 2) Convert to a zero-centered signed integer;

[0026] 3) After normalization, multiply by the reference voltage vref to calculate the actual differential voltage;

[0027] 4) Superimpose a fixed offset to output the absolute voltage value at the IN+ terminal;

[0028] The calibrated data is sent to the terminal computing unit via a custom Bluetooth protocol.

[0029] To address the aforementioned technical problems, this invention also proposes a multi-channel surface electromyography (EMG) information acquisition and processing system for rehabilitation training and assessment of stroke patients, comprising:

[0030] A flexible FPC electrode with three acquisition points and silver chloride paste printed on a steel mesh is used to attach to the surface of the patient's muscle tissue to acquire raw electromyographic signals (sub-electrode signals (IN+, IN) and reference electrode signal (REF)).

[0031] The analog signal processing circuit includes a three-pole active bandpass filter, an RFI filter, and a three-op-amp instrumentation amplifier connected in sequence, which are used to process raw electromyographic signals and output analog signals.

[0032] The signal conversion circuit, including a reference voltage source and a 24-bit ADC, is used to convert analog signals into digital signals;

[0033] The wireless transmission unit transmits digital signals to the terminal computing unit via a custom Bluetooth protocol.

[0034] Furthermore, the three-sampling-point stencil-printed silver chloride paste flexible FPC electrode includes:

[0035] FPC circuit boards serve as signal transmission substrates;

[0036] A silica adhesive layer is applied to the surface of the FPC circuit board for structural bonding and insulation.

[0037] A carbon-silver conductive layer is printed on a silicon dioxide adhesive layer to form a serpentine grid electrode structure;

[0038] The PI reinforcement layer is located in the electrode end connection area, providing mechanical strength and a standardized interface.

[0039] Furthermore, the carbon silver conductive layer is connected to the pads of the FPC circuit board through the openings in the silicon dioxide adhesive layer, forming a wireless integrated connection; the serpentine grid electrode structure is composed of repeated S-shaped or three-arm spiral units laid flat.

[0040] Furthermore, the three operational amplifier instrumentation amplifier consists of three operational amplifiers. The first stage is an instrumentation amplifier that achieves high common-mode rejection ratio and primary filtering. The second stage is a second-order active filter, which together form a three-pole bandpass characteristic.

[0041] Furthermore, the data processing method for the 24-bit ADC includes:

[0042] Use the value representing the negative full scale in the offset binary code of the ADC output as the zero point;

[0043] Subtract the zero-point offset and convert to a 0-centered signed integer;

[0044] After normalization, multiply by the reference voltage vref to calculate the actual differential voltage between IN+ and IN;

[0045] A fixed offset is superimposed to convert it into the absolute voltage value of the IN+ terminal.

[0046] A device for acquiring and processing surface electromyography (EMG) information for rehabilitation training and assessment of stroke patients, used to implement the EMG signal processing method for stroke patient rehabilitation training as described above, comprising:

[0047] A housing having a cavity formed within it;

[0048] An analog-to-digital conversion circuit and a wireless transmission circuit board are disposed within the cavity;

[0049] An analog signal processing amplification and filtering circuit board is disposed in the cavity and electrically connected to the analog-to-digital conversion circuit and the wireless transmission circuit board.

[0050] The three-slot stencil printed silver chloride paste flexible FPC electrode is printed on the silica adhesive layer of the analog signal processing amplification and filtering circuit board, and the three-slot stencil printed silver chloride paste flexible FPC electrode protrudes from the outer surface of the housing on the side away from the analog signal processing amplification and filtering circuit board.

[0051] The three-sampling-point stencil printed silver chloride paste flexible FPC electrode includes an FPC circuit board, a carbon silver conductive layer, and a PI reinforcing layer. The carbon silver conductive layer is printed on the FPC circuit board through a silicon dioxide adhesive layer. The carbon silver conductive layer has a circular serpentine mesh structure, which is formed by multiple curved "S"-shaped or three-arm spiral units connected and laid flat. The PI reinforcing layer is set at the connection area at the end of the electrode to provide mechanical strength and a standardized interface.

[0052] To mitigate the impact of ECG signals, the technical solution of this invention incorporates a front-end dual-pole active filter in the analog circuit section, specifically designed to filter out ECG signals in the 1-10Hz range. To address motion artifacts caused by friction between metal and skin or skin deformation, a silica adhesive layer is added to the FPC electrodes, ensuring good adhesion between the electrodes and the skin while guaranteeing user safety. To combat RF interference, an RF filter is added at the circuit front end. Furthermore, to reduce motion interference caused by wires, this system employs an FPC gold finger connection method, directly connecting to the circuit board, thus avoiding interference that may arise from traditional wire connections. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0054] Figure 1 This is a flowchart of the steps of an electromyographic signal processing method for rehabilitation training of stroke patients according to the present invention.

[0055] Figure 2 This is a schematic diagram of the electromyographic signal processing system for rehabilitation training of stroke patients according to the present invention.

[0056] Figure 3 This is a schematic diagram of the structure of the flexible FPC electrode with silver chloride paste printed on a stencil at three collection points according to the present invention;

[0057] Figure 4 This is a schematic diagram of the analog circuit and digital-to-analog conversion circuit of the present invention;

[0058] Figure 5 This is a simulation diagram of the connection line between the FPC and the gold finger in this invention;

[0059] Figure 6 This is a circuit schematic diagram of the analog signal processing circuit of the present invention;

[0060] Figure 7 This is a time-domain waveform comparison diagram of the normalized electromyography (EMG) signal acquired by the flexible electrode of this invention and the normalized EMG signal acquired by the steel electrode.

[0061] Figure 8 This is a power spectral density spectrum diagram of the normalized electromyography signal acquired by the flexible electrode of the present invention and the normalized electromyography signal acquired by the steel electrode.

[0062] Figure 9This is a single-sided amplitude spectrum diagram of the normalized electromyography signal acquired by the flexible electrode of this invention and the normalized electromyography signal acquired by the steel electrode.

[0063] Figure 10 This is the overall flowchart of the multi-channel surface electromyography information acquisition and processing system for rehabilitation training and assessment of stroke patients according to the present invention.

[0064] Explanation of icon numbers:

[0065] a is the skin; b is the flexible FPC electrode with silver chloride paste printed on a stencil at three acquisition points; c is the analog signal processing circuit; d is the signal conversion circuit and wireless transmission unit; g is the host computer terminal with a Linux operating system; 1. FPC circuit board; 2. Silicon dioxide adhesive layer; 3. Carbon silver conductive layer; 4. PI reinforcement layer. Detailed Implementation

[0066] This invention proposes a method, system, and device for electromyographic signal processing in rehabilitation training for stroke patients, aiming to limit the signal frequency band to the main energy range of electromyographic signals and filter out high-frequency noise and motion artifacts other than muscle signals.

[0067] The electromyographic signal processing method for rehabilitation training of stroke patients proposed in this invention will be described below in specific embodiments:

[0068] Example 1:

[0069] An electromyographic signal processing method for rehabilitation training of stroke patients, such as Figure 1 As shown, it includes the following steps:

[0070] S1. Signal Acquisition and RF Filtering:

[0071] Differential electrode signals (IN+, IN) and reference electrode signals (REF) are acquired using a three-point flexible FPC electrode.

[0072] The three signals are input into the RFI radio frequency filter to filter out high-frequency radio frequency interference;

[0073] S2. Differential Amplification and Three-Stage Filtering:

[0074] The filtered differential signals (IN+, IN) are input into a three-op-amp instrumentation amplifier:

[0075] First stage: The instrumentation amplifier structure is adopted to amplify the differential signal and suppress common-mode interference, while realizing the first stage of low-pass filtering;

[0076] Second stage: High-order filtering is performed through a second-order active filter, which together with the first stage forms a three-pole bandpass characteristic;

[0077] Amplify the output signal amplitude to 2 / 3 of the maximum value of the ADC range;

[0078] S3. Signal boosting and ADC conversion:

[0079] Process the reference electrode signal (REF) to generate a DC bias voltage;

[0080] The DC bias voltage is added to the amplified differential signal to stabilize the bipolar electromyographic signal to the input range of the ADC.

[0081] Analog signals are converted into digital signals using a 24-bit ADC signal converter;

[0082] S4. Data Calibration and Transmission:

[0083] Perform a calibration algorithm on the ADC output data:

[0084] 1) Set the encoded value representing the negative full scale to zero;

[0085] 2) Convert to a zero-centered signed integer;

[0086] 3) After normalization, multiply by the reference voltage vref to calculate the actual differential voltage;

[0087] 4) Superimpose a fixed offset to output the absolute voltage value at the IN+ terminal;

[0088] The calibrated data is sent to the terminal computing unit via a custom Bluetooth protocol.

[0089] The 24-bit ADC signal converter employs special isolation and voltage calculations between itself and the pre-processing circuitry to ensure maximum ADC accuracy. Simultaneously, the converted digital signal is isolated from the Bluetooth chip via a digital isolator. Specifically, a special value representing the negative full-scale range is first treated as a zero point—a fault-tolerant mechanism. Next, by subtracting the zero-point offset, the ADC's offset binary code is converted into a 0-centered signed integer, making it suitable for standard mathematical operations. This signed integer is normalized by dividing by the ADC's full-scale range, resulting in a floating-point number representing the voltage ratio. This floating-point number is then multiplied by the reference voltage vref to accurately calculate the actual differential voltage between the two ADC input pins, IN+ and IN-. Finally, a fixed value is added to the calculated differential voltage. Because the circuit is specifically designed, this addition operation converts the differential measurement result into the absolute voltage value at the positive input terminal (V_IN+).

[0090] Example 2:

[0091] A multi-channel surface electromyography (EMG) information acquisition and processing system for rehabilitation training and assessment of stroke patients includes:

[0092] A flexible FPC electrode with three acquisition points and silver chloride paste printed on a steel mesh is used to attach to the surface of the patient's muscle tissue to acquire raw electromyographic signals (sub-electrode signals (IN+, IN) and reference electrode signal (REF)).

[0093] The analog signal processing circuit includes a three-pole active bandpass filter, an RFI filter, and a three-op-amp instrumentation amplifier connected in sequence, which are used to process raw electromyographic signals and output analog signals.

[0094] The signal conversion circuit, including a reference voltage source and a 24-bit ADC, is used to convert analog signals into digital signals;

[0095] The wireless transmission unit transmits digital signals to the terminal computing unit via a custom Bluetooth protocol.

[0096] Specifically, specifically as Figure 10 As shown, the flexible FPC electrode section with silver chloride paste printed on the stencil at three acquisition points serves as the signal acquisition device, while the analog signal processing, amplification, filtering, and signal conversion circuits serve as the signal hardware processing and amplification. The terminal computing unit with a Linux operating system is used for signal calculation, processing, display, and storage.

[0097] The three-point stencil-printed silver chloride paste flexible FPC electrode consists of an FPC circuit board, a silica adhesive layer, a carbon-silver conductive layer, and a PI reinforcing layer. During signal acquisition, the patient needs to attach the electrode to the muscle tissue area to be acquired, ensuring a tight fit and improving signal acquisition accuracy. The flexible electrode exhibits a higher power spectral density than the rigid electrode, indicating superior external interference suppression capabilities. It effectively isolates and captures pure physiological signal energy, providing higher signal-to-noise ratio and more reliable raw data for subsequent electromyography (EMG) analysis. The flexible electrode also demonstrates better single-sided amplitude spectrum acquisition than the rigid electrode, indicating higher sensitivity and better signal purity in capturing genuine physiological EMG signals. In contrast, the spectrum of the steel electrode is dominated by non-physiological peaks, and the amplitude of its effective physiological signal is weak and submerged in background noise, making effective resolution difficult. In summary, the flexible electrode provides more reliable and easier-to-analyze bioelectrical data.

[0098] After being processed by circuitry and converted from digital to analog, the data is transmitted wirelessly to a terminal computing unit running a Linux operating system. This unit then processes the data using algorithms for calculation, decoding, filtering, classification, storage, and display. The data is stored to analyze the rehabilitation progress, stage the rehabilitation process, and predict the next rehabilitation training intensity. A training report is then generated and provided to the patient or physician.

[0099] The structure is divided into three layers, such as Figure 2As shown, the overall structure is divided into three layers:

[0100] (1) b is the innermost layer that is in close contact with the skin and is used for flexible electrode (three-sampling point steel mesh printed silver chloride paste flexible FPC electrode) collection. It can be perfectly attached to human skin a and collect muscle electrical signals with rehabilitation exercises, making the collected electrical signals more accurate.

[0101] (2) c is the second-layer analog signal processing amplification and filtering circuit (analog signal processing circuit). In signal processing, an RFI filter and a three-pole active filter are used, and a three-stage amplification circuit is used to ensure both signal purity and amplification factor. For a single channel of electromyography signal, a total of three electrodes, a pair of differential electrodes and a reference electrode, are required for acquisition. This can reduce interference and ensure the correctness of acquisition by the subsequent circuit.

[0102] (3)d is the third layer of digital-to-analog conversion circuit and wireless transmission circuit (signal conversion circuit and wireless transmission unit). This digital-to-analog conversion circuit can accurately convert the pre-processed and amplified signal into a digital signal, and then the wireless transmission circuit transmits the data to our data processing terminal g. It has the characteristics of accuracy, speed and stability and can meet the needs of different sampling rates.

[0103] Furthermore, the three-sampling-point stencil-printed silver chloride paste flexible FPC electrode, such as... Figure 3 As shown, it includes:

[0104] FPC circuit board 1 serves as a signal transmission substrate;

[0105] Silicon dioxide adhesive layer 2 is applied to the surface of the FPC circuit board for structural bonding and insulation.

[0106] A carbon-silver conductive layer 3 is printed on a silicon dioxide adhesive layer to form a serpentine grid electrode structure;

[0107] PI reinforcement layer 4 is set in the electrode end connection area to provide mechanical strength and standardized interface.

[0108] Specifically, the PI reinforcement layer provides mechanical strength and a standard interface. It is adhered to the end of the flexible circuit, giving it sufficient rigidity and thickness to ensure stable insertion into the standard connectors of external devices.

[0109] The FPC circuit board is a flexible signal transmission network. It serves as the substrate and "skeleton" of the entire sensor, carrying the copper foil traces from the electrodes to the connectors. Its flexibility allows the sensor to conform to the contours of the body.

[0110] In this structure, the silica adhesive layer plays a crucial role in bonding and insulating the internal structure. It is a very thin dielectric layer (insulating layer) responsible for firmly securing the subsequently printed or laminated carbon-silver conductive layer to the FPC circuit board. Its insulating properties are critical because they ensure that the carbon-silver layer conducts electricity to the FPC's circuitry only in designated locations.

[0111] The carbon-silver conductive layer is the sensor's direct-contact electrode. It is printed on a silica adhesive layer and serves as the outermost layer of the entire sensor, directly contacting the user's skin. It is responsible for capturing the weak biopotentials generated by muscle activity.

[0112] The core advantage of this combination method lies in the direct integration of dry electrodes. Traditional dry electrodes (such as metal sheets) need to be connected to the circuit board through riveting or conductive adhesive, which is complex and has low reliability. This structure, however, directly prints / fixes functional carbon-silver electrode layers onto the flexible circuit through an adhesive layer, forming a highly integrated and seamless dry electrode system. It also embodies the characteristics of combining rigidity and flexibility. The sensor body achieves flexibility through FPC, allowing it to conform to the skin; while the connector end ensures rigidity through a PI reinforcement layer, facilitating connection. This internally bonded structure makes the mechanical properties of the entire flexible part more uniform and stable.

[0113] Another synergistic effect of this design is the simplification of the manufacturing process and the reduction of costs. By combining bonding with electrode fabrication, large-area electrode arrays can be fabricated and fixed in one go using technologies such as screen printing, making it very suitable for mass production and more cost-effective than traditional electrode manufacturing methods.

[0114] Its core principle is "direct lamination integration of functional materials". Instead of relying on an additional skin-friendly conductive adhesive layer, it directly integrates the carbon-silver material with bioelectric signal pickup function onto the FPC signal transmission substrate through a structural silica adhesive dielectric layer.

[0115] The final result is a pure, reusable dry electrode sensor. Because it lacks a hydrogel layer, this sensor requires no conductive paste before use, is easy to clean afterward, and can be reused multiple times, significantly extending its lifespan. While the signal quality of dry electrodes is more sensitive to skin contact, its convenience and reusability are undeniable advantages.

[0116] Furthermore, the carbon silver conductive layer is connected to the pads of the FPC circuit board through the openings in the silicon dioxide adhesive layer, forming a wireless integrated connection; the serpentine grid electrode structure is composed of repeated S-shaped or three-arm spiral units laid flat.

[0117] Furthermore, the three operational amplifier instrumentation amplifier consists of three operational amplifiers. The first stage is an instrumentation amplifier that achieves high common-mode rejection ratio and primary filtering. The second stage is a second-order active filter, which together form a three-pole bandpass characteristic.

[0118] like Figure 4 The diagram shows the flow chart of the analog circuit and the digital-to-analog converter circuit. The signal is transmitted from the electrode to the analog circuit. First, it passes through the RF filter to filter out radio frequency interference. Then, it passes through the amplification and filtering circuit to obtain a pure electromyography signal to avoid interference from power frequency and electrocardiogram signals. Finally, the analog signal is converted into a digital signal by the 24-bit ADC signal converter for transmission to the next stage circuit and sent to the terminal.

[0119] Furthermore, the data processing method for the 24-bit ADC includes:

[0120] Use the value representing the negative full scale in the offset binary code of the ADC output as the zero point;

[0121] Subtract the zero-point offset and convert to a 0-centered signed integer;

[0122] After normalization, multiply by the reference voltage vref to calculate the actual differential voltage between IN+ and IN;

[0123] A fixed offset is superimposed to convert it into the absolute voltage value of the IN+ terminal.

[0124] Figure 5This diagram illustrates the wiring of the FPC electrodes and interface. To prevent the internal metal wires from breaking when the electrodes deform with muscle tissue, and to address issues such as insufficient flexibility leading to a suspended center on the electrode, the circuit board needs to be covered with numerous winding, fine wires. To prevent signal interference introduced by the suspended wires, the electrode leads on the circuit board are designed as standard FPC connectors with gold fingers and a PI reinforcement layer is added to ensure stable connections. When the carbon silver conductive layer is printed on the silicon dioxide adhesive layer, it forms a circular serpentine grid structure. This grid is not a simple square or line, but rather composed of numerous repeating, curved "S"-shaped or three-armed spiral units connected and laid flat. This intricate, fractal-like pattern covers the entire circular electrode area. This unique design is not for aesthetics; each structure is designed to achieve a specific performance effect. The serpentine grid structure is the core of the entire design, its main purpose being to improve the sensor's mechanical compliance. When the sensor is attached to the skin, muscle contraction and relaxation cause skin deformation. If the electrode is a solid metal sheet, repeated stretching and bending will cause stress concentration, which can easily lead to metal fatigue, fracture, or delamination from the substrate, resulting in sensor failure. However, this serpentine mesh structure, with its curved lines acting like countless tiny springs, can absorb and disperse stress through its own stretching and compression, thus ensuring that the electrode itself will not be damaged when the skin deforms, and can always maintain good electrical contact with the skin.

[0125] Three-dimensional structure: The three-dimensional structure of this flexible electromyography (EMG) sensor is constructed using a multi-layer composite process. The bottom layer is a flexible flexible printed circuit board (FPC), serving as the substrate for the entire sensor, upon which internal copper foil circuitry is laid. Above the FPC, a silica adhesive layer acts as a structural adhesive, but with precise openings in the via connection areas to expose the underlying connection pads. The top layer, the functional layer that directly contacts the skin, is a carbon-silver conductive layer, printed or fixed onto the adhesive layer, forming the serpentine grid electrode pattern shown in the figure. Through the openings in the adhesive layer, the carbon-silver layer achieves electrical conductivity with the pads on the FPC, thereby introducing the acquired signal into the internal circuitry. Finally, the PI reinforcement layer is not located in the electrode sensing area but is adhered to the back of the end of the entire flexible circuit strip (connector end) to increase the rigidity and thickness of this area, facilitating a reliable physical connection with external devices.

[0126] Figure 6 The circuit diagram of this invention is shown. The electromyography signal acquisition consists of a pair of differential electrodes and a reference electrode. First, these three sets of signals are filtered out by an RF radio frequency filter to remove radio frequency interference. Then, the differential signal is processed and amplified to 2 / 3 of the maximum value of the ADC range by two specially designed three-pole filter amplifiers. The reference electrode signal is amplified and added to the differential signal to achieve the effect of signal boosting.

[0127] This circuit is a well-designed, high-precision bioelectrical signal conditioning link. Its core task is to effectively filter and amplify the weak and easily interfered electromyographic signals collected from the human body surface, ultimately converting them into stable signals suitable for sampling by an analog-to-digital converter (ADC). Its working principle can be summarized as a series processing procedure of "front-end interference suppression, mid-end amplification and filtering, and back-end level boosting and fine filtering".

[0128] Specifically: First, three sets of signals, consisting of a pair of differential electrodes and a reference electrode, enter the front end of the circuit. The first stop for these signals is the RFI (Radio Frequency Identifier) ​​filter, a crucial protective element. Since the human body and wires act like antennas, they easily pick up high-frequency radio frequency interference from the environment. If this interference directly enters the amplifier, it will be rectified into an incorrect DC bias. The RFI filter, as a pre-amplifier low-pass filter, effectively filters out these harmful high-frequency components, ensuring that only relatively pure bioelectrical signals enter the core processing circuit. This is fundamental to guaranteeing the measurement accuracy of the entire system.

[0129] Next, the pre-filtered differential signal is fed into the instrumentation amplifier stage, which consists of the first two operational amplifiers. This stage is the heart of the entire circuit; it utilizes a cross-coupled structure to achieve an extremely high common-mode rejection ratio (CMRR), accurately amplifying the difference between the two input signals (i.e., the true electromyographic signal) while vigorously suppressing common interferences on both signals (such as 50 / 60Hz power frequency noise). While amplifying, it also provides a low-pass filter with a very steep roll-off characteristic, effectively limiting the signal bandwidth to the main energy range of the electromyographic signal and filtering out high-frequency noise and motion artifacts other than those related to muscle signals.

[0130] The amplified signal then enters a second-order active filter composed of a third operational amplifier. It is cascaded with the filtering function of the previous stage and together they form the higher-order part of the "three-pole filter amplifier", which further enhances the ability to suppress out-of-band noise.

[0131] Finally, the signal from the reference electrode (REF), after processing, is introduced into the instrumentation amplifier or subsequent circuitry. Its core function is to superimpose the amplified and filtered AC electromyography (EMG) signal onto a stable DC bias voltage. This elevates the bipolar signal, which would otherwise fluctuate around zero, into a positive voltage range. This is crucial for single-supply ADCs, ensuring that the negative half-cycle of the signal is not clipped, thus preserving all the information of the EMG signal and adjusting its amplitude to approximately two-thirds of the ADC's range for optimal quantization resolution and dynamic range. Finally, the signal may pass through a passive RC filter before output for final refinement, ensuring its stable acquisition by the subsequent ADC stage.

[0132] Figure 7 This image shows a comparison of the time-domain waveforms of normalized electromyographic (EMG) signals acquired by the flexible electrode and those acquired by the steel electrode. In the comparison of the time-domain waveforms of the normalized EMG signals, the flexible electrode exhibits a significant signal quality advantage over the steel electrode. The zero-center signal alignment characteristic of the flexible electrode not only provides a more intuitive and clear representation of muscle activity, making the distinction between muscle contraction and relaxation states more explicit, but also greatly optimizes the effective resolution of signal amplitude changes, thereby enabling more faithful and sensitive capture and reflection of real physiological EMG activity.

[0133] Figure 8 The images show the power spectral density (PSD) spectra of normalized electromyography (EMG) signals acquired by the flexible electrode and the steel electrode, respectively. Comparing the PSD spectra of the steel electrode (left) and the flexible electrode (right), the flexible electrode exhibits significantly superior signal quality. The PSD of the flexible electrode clearly shows the broadband energy distribution characteristic of physiological EMG signals, specifically a distinct "hump"-shaped energy concentration in the low-frequency range (approximately 0-150 Hz), followed by a smooth energy decay, faithfully reflecting the frequency characteristics of muscle activity. In stark contrast, the PSD spectra of the steel electrode not only exhibit significant, non-physiological discrete interference peaks at specific frequency points (approximately 50 Hz in the figure), but its overall energy distribution is also relatively dispersed, masking the true physiological EMG characteristics.

[0134] Figure 9 This is a comparison of the normalized electromyography (EMG) signals acquired by the flexible electrode of this invention and those acquired by the steel electrode, showing their single-sided amplitude spectra. Comparing the single-sided amplitude spectra of the steel electrode (left) and the flexible electrode (right), the flexible electrode exhibits significant advantages in signal quality and practicality. The spectrum of the flexible electrode clearly shows the inherent broadband energy distribution of physiological EMG signals, namely, concentration in the low-frequency range followed by smooth decay, without any dominant, non-physiological discrete frequency peaks. This directly indicates its excellent ability to suppress external interference. Furthermore, the physiological signal amplitude captured by the flexible electrode is significantly higher, with a Y-axis scale of x 10. -3 The steel electrode is only x 10 -4 This fully demonstrates that flexible electrodes have higher sensitivity and better signal purity in capturing real physiological electromyographic signals.

[0135] Table 1 is a comparison of the properties of the silver / silver chloride electrode material used in this invention with various other electrode materials. The selection of electrode materials in electrophysiological signal acquisition is crucial for signal quality and stability. Different electrode materials have varying impedance ranges, stability, electrochemical characteristics, and impacts on signal quality, making them suitable for different applications. Silver / silver chloride electrodes possess extremely low and stable impedance (below several thousand ohms) and high electrochemical stability. Their non-polarizing properties give them high reversibility, providing optimal signal quality with minimal noise and artifacts, and they are commonly used for high-precision biosignal acquisition. Silver / silver chloride and gold electrodes exhibit excellent signal quality and stability, making them suitable for applications requiring high signal accuracy; while stainless steel and carbon-based electrodes are suitable for applications with high comfort requirements or high noise tolerance.

[0136] Table 1

[0137]

[0138] Example 3:

[0139] A device for acquiring and processing surface electromyography (EMG) information for rehabilitation training and assessment of stroke patients, used to implement the EMG signal processing method for stroke patient rehabilitation training as described in Example 1, comprising:

[0140] A housing having a cavity formed within it;

[0141] An analog-to-digital conversion circuit and a wireless transmission circuit board are disposed within the cavity;

[0142] An analog signal processing amplification and filtering circuit board is disposed in the cavity and electrically connected to the analog-to-digital conversion circuit and the wireless transmission circuit board.

[0143] The three-slot stencil printed silver chloride paste flexible FPC electrode is printed on the silica adhesive layer of the analog signal processing amplification and filtering circuit board, and the three-slot stencil printed silver chloride paste flexible FPC electrode protrudes from the outer surface of the housing on the side away from the analog signal processing amplification and filtering circuit board.

[0144] The three-sampling-point stencil printed silver chloride paste flexible FPC electrode includes an FPC circuit board, a carbon silver conductive layer, and a PI reinforcing layer. The carbon silver conductive layer is printed on the FPC circuit board through a silicon dioxide adhesive layer. The carbon silver conductive layer has a circular serpentine mesh structure, which is formed by multiple curved "S"-shaped or three-arm spiral units connected and laid flat. The PI reinforcing layer is set at the connection area at the end of the electrode to provide mechanical strength and a standardized interface.

[0145] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for processing electromyographic signals for rehabilitation training of stroke patients, characterized in that, Includes the following steps: S1. Signal Acquisition and RF Filtering: Differential electrode signals and reference electrode signals are acquired using a three-point flexible FPC electrode. The three signals are input into the RFI radio frequency filter to filter out high-frequency radio frequency interference; S2. Differential Amplification and Three-Stage Filtering: The filtered differential electrode signal is then input into a three-op-amp instrumentation amplifier: First stage: The instrumentation amplifier structure is adopted to amplify the differential electrode signal and suppress common-mode interference, while realizing the first stage of low-pass filtering; Second stage: High-order filtering is performed through a second-order active filter, which together with the first stage forms a three-pole bandpass characteristic; Amplify the output signal amplitude to 2 / 3 of the maximum value of the ADC range; S3. Signal boosting and ADC conversion: Process the reference electrode signal to generate a DC bias voltage; The DC bias voltage is added to the amplified differential electrode signal to stabilize the bipolar electromyographic signal to the input range of the ADC. Analog signals are converted into digital signals using a 24-bit ADC signal converter; S4. Data Calibration and Transmission: Perform a calibration algorithm on the ADC output data: Set the encoded value representing the negative full scale to zero; Convert to a zero-centered signed integer; After normalization, multiply by the reference voltage vref to calculate the actual differential voltage; By superimposing a fixed offset, the absolute voltage value of the IN+ terminal is output; The calibrated data is sent to the terminal computing unit via a custom Bluetooth protocol.

2. A multi-channel surface electromyography (EMG) information acquisition and processing system for rehabilitation training and assessment of stroke patients, used to execute the EMG signal processing method for rehabilitation training of stroke patients as described in claim 1, characterized in that, include: Three-point stencil printed silver chloride paste flexible FPC electrodes are used to attach to the surface of the patient's muscle tissue to collect raw electromyographic signals; The analog signal processing circuit includes a three-pole active bandpass filter, an RFI filter, and a three-op-amp instrumentation amplifier connected in sequence, which are used to process raw electromyographic signals and output analog signals. The signal conversion circuit, including a reference voltage source and a 24-bit ADC, is used to convert analog signals into digital signals; The wireless transmission unit transmits digital signals to the terminal computing unit via a custom Bluetooth protocol.

3. The multi-channel surface electromyography information acquisition and processing system for rehabilitation training and assessment of stroke patients as described in claim 2, characterized in that: The three-sampling-point stencil-printed silver chloride paste flexible FPC electrode includes: FPC circuit boards serve as signal transmission substrates; A silica adhesive layer is applied to the surface of the FPC circuit board for structural bonding and insulation. A carbon-silver conductive layer is printed on a silicon dioxide adhesive layer to form a serpentine grid electrode structure; The PI reinforcement layer is located in the electrode end connection area, providing mechanical strength and a standardized interface.

4. The multi-channel surface electromyography information acquisition and processing system for rehabilitation training and assessment of stroke patients as described in claim 3, characterized in that: The carbon-silver conductive layer is connected to the circuit board pads through the openings in the silicon dioxide adhesive layer, forming a wireless integrated connection; the serpentine grid electrode structure is composed of repeated S-shaped or three-arm spiral units laid flat.

5. The multi-channel surface electromyography information acquisition and processing system for rehabilitation training and assessment of stroke patients as described in claim 2, characterized in that: The three operational amplifier instrumentation amplifier consists of three operational amplifiers. The first stage is an instrumentation amplifier that achieves high common-mode rejection ratio and primary filtering. The second stage is a second-order active filter, which together form a three-pole bandpass characteristic.

6. The multi-channel surface electromyography information acquisition and processing system for rehabilitation training and assessment of stroke patients as described in claim 2, characterized in that: The data processing method for the 24-bit ADC includes: Use the value representing the negative full scale in the offset binary code of the ADC output as the zero point; Subtract the zero-point offset and convert to a 0-centered signed integer; After normalization, multiply by the reference voltage vref to calculate the actual differential voltage between IN+ and IN; A fixed offset is superimposed to convert it into the absolute voltage value of the IN+ terminal.

7. A device for acquiring and processing multi-channel surface electromyography (EMG) information for rehabilitation training and assessment of stroke patients, used to implement the EMG signal processing method for rehabilitation training of stroke patients as described in claim 1, characterized in that, include: A housing having a cavity formed within it; An analog-to-digital conversion circuit and a wireless transmission circuit board are disposed within the cavity; An analog signal processing amplification and filtering circuit board is disposed in the cavity and electrically connected to the analog-to-digital conversion circuit and the wireless transmission circuit board. The three-slot stencil printed silver chloride paste flexible FPC electrode is printed on the silica adhesive layer of the analog signal processing amplification and filtering circuit board, and the three-slot stencil printed silver chloride paste flexible FPC electrode protrudes from the outer surface of the housing on the side away from the analog signal processing amplification and filtering circuit board. The three-sampling-point stencil printed silver chloride paste flexible FPC electrode includes an FPC circuit board, a carbon silver conductive layer, and a PI reinforcing layer. The carbon silver conductive layer is printed on the FPC circuit board through a silicon dioxide adhesive layer. The carbon silver conductive layer has a circular serpentine mesh structure, which is formed by multiple curved "S"-shaped or three-arm spiral units connected and laid flat. The PI reinforcing layer is set at the connection area at the end of the electrode to provide mechanical strength and a standardized interface.

Images