Muscle status monitoring method and sensor

Abstract

The application provides a muscle state monitoring method and a sensor, the method adopts a sensor which can be directly attached to the skin surface of the muscle, and collects a voltage signal generated due to muscle deformation. Through a set of effective wave peak identification rules, the number of wave peaks, peak-to-peak values and waveform characteristics in the signal are analyzed, and the direct detection of the muscle micro-force state is realized. Further, the application also provides a method for realizing gesture recognition, muscle strength limit determination and muscle fatigue degree grading based on the same sensor. The application solves the problems in the prior art, such as indirect detection, single function dimension, lack of quantitative rules for signal analysis and the like, and has the advantages of direct and accurate detection, multi-dimensional integrated evaluation, strong anti-interference ability, comfortable wearing and no hindrance to movement.

Description

Technical Field

[0001] This invention relates to the field of physiological signal monitoring technology, specifically to a method and sensor for monitoring muscle state. Background Technology

[0002] Precise monitoring of muscle exertion and fatigue levels has significant application value in scenarios such as stroke patient rehabilitation training, scientific training for athletes, postoperative functional recovery of hand injuries, and sign language recognition-assisted communication. In particular, as the most delicate part of the human body, the state of muscle control in the hand is a key indicator for assessing motor function, training effectiveness, and fatigue risk.

[0003] Currently, the mainstream muscle condition monitoring technologies mainly fall into the following two categories: The first category is electromyography (EMG) sensor technology. This method indirectly assesses muscle activity by collecting bioelectrical signals generated by the superposition of action potentials of motor units during muscle contraction through electrodes attached to the skin surface. However, this technology has inherent drawbacks: the relationship between EMG signals and muscle contraction force is non-linear, making it difficult to accurately quantify muscle strength; signal acquisition is susceptible to environmental electromagnetic interference, motion artifacts, and crosstalk from neighboring muscle EMG signals, resulting in a low signal-to-noise ratio and difficulties in signal analysis.

[0004] The second category is based on flexible strain or piezoresistive sensors. These sensors are typically attached to the skin surface of joints (such as finger joints). Their working principle is that when the joint bends, the sensor attached to the joint stretches or compresses along with the skin, thus outputting an electrical signal. A significant problem with this method is the indirectness of detection: it measures the skin deformation caused by joint bending, rather than the muscle's own exertion. Since joint deformation and muscle exertion are not strictly correlated, this indirect measurement method introduces systematic errors. Furthermore, the sensor's attachment to the joint inevitably hinders normal joint movement to some extent, affecting the wearer's natural movement. In addition, the sensitivity and resolution of existing sensors of this type are generally insufficient, mainly suitable for detecting large-amplitude joint movements, and difficult to capture subtle muscle activity signals such as slight bending of a single finger or the faint trembling at the beginning of muscle exertion.

[0005] Therefore, existing technologies cannot provide a muscle monitoring solution that is high-resolution, easy to wear, does not hinder movement, and can directly capture the true state of muscle exertion. Summary of the Invention

[0006] This invention aims to solve the following technical problems existing in muscle state monitoring technology: First, the detection method is indirect, either based on nonlinear inference of electrophysiological signals or indirect back-inference based on joint deformation, which cannot directly reflect the true physical force exertion state of the muscle; Second, the functional dimension is single, and it is impossible to achieve multi-dimensional assessment from micro-force identification to fatigue grading through the same device and process; Third, there is a lack of quantitative rules for analyzing sensor output signals, especially for the identification of effective peaks, the differentiation of independent movements, and the definition of fatigue levels, which lacks engineering-implementable criteria.

[0007] To achieve the above objectives, the present invention provides a method for monitoring muscle condition, comprising the following steps: A flexible piezoelectric sensor is directly attached to the skin surface of the target muscle, ensuring close contact between the sensor and the skin. The voltage signal output by the sensor due to sensing muscle deformation is collected. The collected voltage signal is analyzed according to a preset effective peak identification rule. Based on the number of peaks, peak value, and / or waveform change characteristics in the signal, the muscle's force exertion state is determined.

[0008] Furthermore, the rules for identifying valid peaks include: a peak with a distinct peak, forming a sinusoidal signal and being greater than the noise level, is considered a valid peak.

[0009] Furthermore, determining the muscle's exertion state includes detecting minute exertion states: when a single independent peak is detected, it is determined to be an independent movement of a single joint; when multiple independent peaks are detected superimposed, it is determined to be a coordinated movement of multiple joints; among them, the joint bending angle is positively correlated with the peak-to-peak value of the corresponding peak.

[0010] Furthermore, it also includes gesture recognition steps: attaching a flexible piezoelectric sensor to the muscle controlling the target gesture; collecting voltage signals when the test subject performs different gestures; and distinguishing different gestures based on the number of effective peaks, peak-to-peak value, and timing of peak appearance in the voltage signal.

[0011] Furthermore, the procedure also includes steps for determining the limits of muscle strength: the test subject holds instruments with different resistance levels in sequence to perform gradient tests; the voltage signals output by the sensors at each resistance level are recorded; when the peak-to-peak value of the voltage signal increases gradually with the increase of the resistance level, and obvious jitter appears on the signal waveform, it is determined that the muscle is close to its strength limit.

[0012] Furthermore, it also includes a muscle fatigue grading step: controlling the test subject to move continuously at a fixed frequency and amplitude; continuously collecting the voltage signal output by the sensor and monitoring the change of the signal peak value over time; and grading the muscle fatigue level according to the degree of decay of the signal peak value relative to the stable amplitude in the initial stage of movement.

[0013] Furthermore, the classification of muscle fatigue includes: using the stable peak-to-peak value of the signal at the initial stage of continuous exercise as a baseline; continuous monitoring, recording the maximum attenuation amount when the peak-to-peak value of the signal decays to a point where it tends to stabilize and the fluctuation is extremely small, which is judged as severe fatigue; during continuous monitoring, when the signal attenuation amount reaches approximately 50% of the maximum attenuation amount, it is judged as moderate fatigue; before reaching moderate fatigue, when the signal shows a continuous and perceptible slight attenuation, it is judged as mild fatigue.

[0014] Furthermore, the flexible piezoelectric sensor is attached to the skin via a flexible wearing component, which includes medical tape and a PU material strip. The sensor is fixed to the PU material strip by the medical tape, and the PU material strip is wrapped and fixed to the target muscle in a pre-stretched state.

[0015] Furthermore, the target muscles are the extensor muscles, flexor carpi radialis, brachioradialis, or frontalis.

[0016] The present invention also provides a sensor comprising a piezoelectric composite film layer, a metal electrode layer, and a silicone encapsulation layer, wherein the piezoelectric composite film layer is formed by dispersing modified piezoelectric particles and polymer materials, followed by film laying, hot pressing, and high-pressure polarization treatment, with a film thickness of 50–80 µm and a sensor resolution greater than 3 µN, for use in muscle mechanical signal detection; its static noise peak-to-peak value is less than 0.1 mV, and its baseline drift is less than 0.2 mV after continuous wear for 1 hour in a pre-stretched state.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This method involves directly attaching the sensor to the skin surface of the target muscle, collecting the physical deformation signals generated during muscle contraction, rather than joint motion or bioelectrical signals. This allows the detected signal to directly and linearly reflect the actual force exertion state of the muscle, fundamentally avoiding systematic errors caused by indirect inference based on joint deformation, as well as the nonlinearity and susceptibility to interference problems of electromyographic signals.

[0018] Based on the same sensor and the same wearing method, multiple functions such as detecting subtle muscle exertion, recognizing gestures, determining muscle strength limits, and grading muscle fatigue can be achieved sequentially or selectively through different analysis processes. This realizes "one method covering multiple needs" and significantly improves the integration and efficiency of monitoring.

[0019] This invention explicitly proposes an "effective peak identification rule," providing a clear and programmable judgment standard for transforming continuous, analog voltage waveforms into discrete, meaningful muscle movement events. Simultaneously, the proposed relative fatigue grading method based on maximum attenuation provides an engineering solution for fatigue assessment that does not rely on absolute amplitude thresholds and exhibits good individual adaptability.

[0020] The sensor is fixed to the muscle belly, rather than at a joint. During movements such as raising an arm, clenching a fist, or running, the sensor can naturally deform with the muscle and skin without being pulled or displaced, and does not affect normal human activity at all. It is suitable for long-term use by rehabilitation patients or athletes. Attached Figure Description

[0021] Figure 1 This is a schematic diagram showing the wearing position of the sensor of the present invention on the extensor muscles of the forearm; Figure 2 This is a diagram showing the output voltage signal of the second finger joint of the right index finger during the movement of the finger micro-joints and bending. Figure 3 This is a diagram showing the output voltage signals of the index finger of the present invention when bent to 30°, 60° and 90° respectively; Figure 4 This is a diagram of the output voltage signal for the gesture recognition test of this invention; Figure 5 This is a schematic diagram showing the decay of the output voltage signal over time during the muscle fatigue test of this invention; Figure 6 This is a comparison diagram of the signals output by the sensor of this invention and the electromyography sensor under the same conditions; Figure 7 This is a verification diagram of the repeatability accuracy of the sensor of this invention after 24,000 repeated tests at a frequency of 20Hz. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0023] The sensor used in this invention comprises a piezoelectric composite film layer, a metal electrode layer, and a silicone encapsulation layer. The piezoelectric composite film layer is formed by dispersing modified piezoelectric particles and polymer materials, followed by film deposition, hot pressing, and high-pressure polarization treatment. The film thickness is 50–80 µm, and the sensor resolution is greater than 3 µN, used for muscle biomechanical signal detection. The static noise peak-to-peak value is less than 0.1 mV, and the baseline drift after continuous wear for 1 hour in a pre-stretched state is less than 0.5 mV. In a 24,000-cycle (20 Hz, 1200 s) repeatability test, its repeatability accuracy reaches 97.4%. Figure 7 As shown.

[0024] Example 1: Detection of minute muscle exertion The sensor is encapsulated in a PU material strip.

[0025] according to Figure 1As shown, the sensor is worn on the extensor digitorum belly of the test subject's right forearm via a PU material band, ensuring that the lower surface of the sensor is in direct and close contact with the skin surface without covering the wrist or finger joints. The sensor electrode output is connected to a data acquisition device.

[0026] First, the baseline signal from the sensor was recorded when there was no muscle activity to confirm that the background noise level was within the normal range. Then, the test subject was instructed to perform the following actions: Action 1: Bend only the first joint of the right index finger, keeping the other joints still. At this moment, the finger extensor muscles undergo a corresponding minute deformation, and the sensor outputs a voltage waveform. For example... Figure 2 As shown in the corresponding part, a distinct peak appears, forming a sinusoidal signal that is greater than the noise, and there are no other peaks before or after that that meet the independent peak determination criteria. Therefore, it is determined to be a single valid peak, corresponding to the independent movement of the first finger joint.

[0027] Action 2: Bend the second joint of your right index finger. (Example:...) Figure 2 As shown in the corresponding section, the output waveform at this time exhibits two peaks, one primary and one secondary. This detection result objectively reflects the physiological structural characteristics of the human body: when bending the second joint of the index finger, a slight movement of the first joint is inevitably caused, resulting in two independent force exertion events. Traditional low-sensitivity sensors cannot capture the minute associated movements of the first joint, while this method, through a high-sensitivity sensor combined with independent peak determination rules, successfully achieves direct detection of such minute and precise force exertion states.

[0028] Specifically, when bending the second joint of the finger, the output waveform contains two peaks: one primary and one secondary. In actual signals, due to the unavoidable physiological tremors during finger joint movement, tiny jitter signals may be superimposed on the signal waveform.

[0029] A valid peak is defined as one that exhibits a distinct peak, forming a sinusoidal signal that is greater than the noise level (i.e., exceeding the noise baseline by 0.1mV). In some waveforms, numerous small peaks may appear due to jitter; these can be ignored. For example, during the bending of the first joint, although there are many small jitters, they can be considered a single peak because the jitter characteristic is present in every waveform. Therefore, this jitter can be disregarded during data processing. In the second joint, the overall waveform is smoother than the first, with fewer fluctuations. This prominent second peak can then be considered a valid peak, and this characteristic is present in every waveform. Therefore, when identifying valid peaks, it is crucial to distinguish between these high-frequency, small-amplitude jitters and sinusoidal peak signals generated by muscle exertion. The specific distinction criteria are: the characteristic peak must exhibit a distinct peak shape relative to the jitter noise, its amplitude must be significantly greater than the amplitude of the jitter signal, and it must show reproducible regularity in repeated tests of the same movement. Only peaks meeting these criteria can be considered valid.

[0030] Action 3: Bend your index finger to a small angle (approximately 30°), a medium angle (approximately 60°), and a large angle (approximately 90°). For example... Figure 2 As shown, the peak-to-peak value of the output waveform increases sequentially with the increase of the bending angle, exhibiting a good positive correlation. This further verifies the reliability and quantizability of the signal detected by this method.

[0031] Example 2: Gesture Recognition The sensor was worn on the finger extensor muscles of the test subject in the same manner as in Example 1. This example demonstrates the recognition of a specific gesture.

[0032] The test subject was instructed to perform the sign language "hello" continuously: first extend the index finger, then raise the thumb. These two sub-movements activated different force patterns of the finger extensor muscles.

[0033] Observe and record the complete voltage waveform output by the sensor, such as Figure 4 As shown, the waveform characteristics are as follows: First, a distinct high-amplitude peak appears, with a relatively large peak-to-peak value, corresponding to the strong contraction of the extensor muscles during the "extending the index finger" action; then, a second peak appears after a short time interval, with a peak-to-peak value significantly lower than the first peak, corresponding to the relatively weaker force exertion of the extensor muscles during the "raising the thumb" action.

[0034] By performing peak detection and feature extraction on the voltage signal, the gesture can be clearly identified as "hello" based on the timing of the two peaks (higher first, then lower) and their amplitude difference (higher amplitude corresponds to an extended index finger, lower amplitude corresponds to a raised thumb). Compared with existing methods that attach sensors to joints to measure deformation angles, this method directly detects muscle force, enabling the recognition of fine gestures that do not rely on large joint flexion. Furthermore, it requires only a single sensor at the muscle, making it more economical and convenient.

[0035] Example 3: Determining the Limits of Muscle Strength The sensor is worn on the radial side of the forearm of the test subject, near the flexor carpi radialis muscle. A set of grip balls with different weights (in this embodiment, different resistance levels are achieved using a set of grip balls with different weights, namely 5kg, 15kg, 25kg, etc.), is prepared as a specific example of resistance gradients. The test subject was instructed to grip each grip ball with maximum force, starting from the smallest kilogram, and the output voltage signal of the sensor was collected and recorded at each maximum force application.

[0036] For the recorded data, the peak-to-peak value of the output signal at each resistance level was calculated. The peak-to-peak values ​​measured at different kilogram weights were plotted as curves or subjected to gradient comparison analysis. In the lower kilogram weight range, the peak-to-peak value showed a clear increasing trend with increasing kilogram weight. When the kilogram weight increased to a certain higher level, the increase in peak-to-peak value became significantly slower, and the amplitude increment between adjacent gradients decreased significantly. Simultaneously, observing the waveform details of the output signal at this higher kilogram weight revealed irregular high-frequency jitter components superimposed on the waveform, with amplitudes significantly increased compared to the previous stage and frequencies higher than the dominant motion frequency. These jitter components exceeded 5% of the peak-to-peak value of the motion wave.

[0037] According to the method of the present invention, the simultaneous occurrence of two phenomena—"signal amplitude growth tending to level off" and "signal exhibiting obvious jitter"—is used as a composite criterion for indicating that the muscle is approaching its strength limit. Based on this, it can be determined that the grip strength at this point is close to the strength limit of the radial wrist flexor muscle in the test subject's forearm. This determination method is more objective and sensitive than simply relying on amplitude saturation or subjective feeling.

[0038] Example 4: Grading of Muscle Fatigue Level The sensor was worn on the extensor muscles of the right forearm of the test subject. An external metronome was set at a frequency of 1Hz, and the test subject was instructed to follow the beat and continuously perform finger flexion and extension movements with a fixed amplitude. The data acquisition device was then activated to record the voltage signal output by the sensor throughout the entire process.

[0039] like Figure 5 As shown, the signal exhibits typical phased attenuation throughout the entire continuous motion process.

[0040] Phase 1 (approximately 0 to 40 seconds): The signal waveform is regular, with the peak-to-peak value of the continuous waves remaining at a relatively stable level, without a significant downward trend. Record the peak-to-peak value of the stable signal during this phase. This phase is considered an indication that the muscles have not experienced significant fatigue.

[0041] Phase Two (approximately 40 to 100 seconds): The peak-to-peak value of the signal begins to show a perceptible, slight, and continuous downward trend, but the rate of decline is slow. This phase is identified as mild fatigue.

[0042] Phase 3 (approximately 100 to 120 seconds): The downward trend of the signal peak-to-peak value becomes significant and accelerates, and the attenuation is quite noticeable. This phase is considered moderate fatigue. At this point, the signal attenuation is approximately half of the final maximum attenuation.

[0043] Phase Four (approximately 120 seconds later): The signal peak-to-peak value drops to a very low level and stabilizes, showing no further significant changes. The waveform becomes irregular and disordered, with numerous high-frequency, small-amplitude fluctuations. At this point, the test subject indicates that it can no longer maintain standard movements. Record the final signal value at this point and calculate the maximum attenuation throughout the entire movement process. This phase is considered severe (complete) fatigue.

[0044] In the method of this invention, the grading of muscle fatigue is based on a relative judgment method with the maximum attenuation amount as a reference: when the attenuation amount reaches approximately 50% of the maximum attenuation amount, it corresponds to moderate fatigue; when the maximum attenuation amount is reached and the signal tends to be flat and disordered, it corresponds to severe fatigue. This method provides a clear and operable engineering assessment tool for muscle fatigue.

[0045] Example 5: Signal-to-noise ratio comparison To verify the signal-to-noise ratio advantage of the sensor based on this invention, a flexible piezoelectric sensor used in this invention and a commercial electromyography sensor were simultaneously attached to the forearm extensor muscles of the same test subject. The test subject was instructed to perform the same finger flexion and extension movements, and both signals were acquired simultaneously.

[0046] The collected signal comparison diagram is as follows Figure 6 As shown, it can be intuitively seen that the signal output by the electromyography (EMG) sensor is mixed with a large amount of background noise and irregular interference spikes, and the motion-related signal features are not prominent. In contrast, the flexible piezoelectric sensor used in this invention has a very clean output background, and the corresponding motion peak shape is clear and its features are significant. The signal-to-noise ratio (SNR) is defined as the ratio of the peak-to-peak value of the motion peak to the root mean square (RMS) value of the background noise during the non-motion period. Under the same test conditions, the SNR of the sensor of this invention is X, while the SNR of the commercial EMG sensor is Y. The former's SNR is improved by more than 90% (XY) / Y×100%=90%. This advantage provides a high-quality source signal guarantee for the effective peak identification and waveform feature analysis steps in the method of this invention.

[0047] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A method for monitoring muscle condition, characterized in that, Includes the following steps: The sensor is directly attached to the skin surface of the target muscle, so that the sensor fits tightly against the skin; The voltage signal output by the sensor due to sensing muscle deformation is collected; The collected voltage signal is analyzed according to the preset effective peak identification rules. Based on the number of peaks, peak-to-peak value and waveform change characteristics in the signal, the muscle exertion state is determined.

2. The muscle state monitoring method according to claim 1, characterized in that, The valid peak identification rules include: a peak with a distinct peak, forming a sinusoidal signal and being greater than the noise level, is considered a valid peak.

3. The muscle state monitoring method according to claim 1, characterized in that, The determination of muscle exertion state includes the detection of minute exertion states: When a single independent peak is detected, it is determined to be an independent movement of a single joint; When multiple independent peaks are detected overlapping, it is determined to be a coordinated movement of multiple joints; This includes a step of determining the joint bending angle based on the peak-to-peak value of the wave crest.

4. The muscle state monitoring method according to claim 1, characterized in that, It also includes gesture recognition steps: The flexible piezoelectric sensor is attached to the muscle that controls the target hand gesture. Collect voltage signals when the test subject performs different hand gestures; Different gestures can be distinguished based on the number of effective peaks, the peak-to-peak value, and the timing of the peaks in the voltage signal.

5. The muscle state monitoring method according to claim 1, characterized in that, It also includes steps for determining muscle strength limits: The test subject was instructed to hold instruments of different resistance levels in sequence to perform gradient tests; Record the voltage signal output by the sensor at each resistance level; When the increase in peak-to-peak value of the voltage signal with increasing resistance level becomes gradual, and obvious jitter appears on the signal waveform, it is determined that the muscle is approaching its strength limit.

6. The muscle state monitoring method according to claim 1, characterized in that, It also includes a muscle fatigue level grading process: The test object is controlled to move continuously at a fixed frequency and amplitude; The voltage signal output by the sensor is continuously acquired, and the change of the peak-to-peak value of the signal over time is monitored. The degree of muscle fatigue is graded based on the degree of attenuation of the signal peak value relative to the stable amplitude in the initial stage of exercise.

7. The muscle state monitoring method according to claim 6, characterized in that, The grading of muscle fatigue includes: The peak-to-peak value of the stable signal during the initial stage of continuous motion is used as the baseline; Continuous monitoring is conducted. When the peak-to-peak value of the signal decays to a point where it tends to stabilize and the fluctuation is extremely small, the maximum decay amount at this point is recorded, and this is determined to be severe fatigue. During the continuous monitoring process, when the signal attenuation reaches 50% of the maximum attenuation, it is determined to be moderate fatigue; Mild fatigue is defined as a continuous and perceptible slight decay of the signal before reaching moderate fatigue.

8. The muscle state monitoring method according to claim 1, characterized in that... The flexible piezoelectric sensor is attached to the skin via a flexible wearing component, which includes medical tape and a PU material strip. The sensor is fixed to the PU material strip by the medical tape, and the PU material strip is wrapped and fixed to the target muscle in a pre-stretched state.

9. The muscle state monitoring method according to claim 1, characterized in that, The target muscles are the extensor muscles, flexor carpi radialis, brachioradialis, or frontalis.

10. A sensor for implementing the muscle state monitoring method as described in any one of claims 1-9, characterized in that, It includes a piezoelectric composite film layer, a metal electrode layer and a silicone encapsulation layer, wherein the piezoelectric composite film layer is formed by dispersing modified piezoelectric particles and polymer materials, followed by film laying, hot pressing and high-pressure polarization treatment.

Images