Sinopenia electromyographic signal acquisition and analysis system
By combining flexible dry electrode arrays and deep learning models, the problems of expensive equipment and subjectivity in existing sarcopenia diagnosis methods have been solved, enabling non-invasive and convenient early screening and accurate diagnosis of sarcopenia, and improving the accuracy of muscle function assessment.
Patent Information
- Application Number
- CN202511988960.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-26
- Publication Date
- 2026-02-17
AI Technical Summary
Existing diagnostic methods for sarcopenia involve expensive equipment, complex procedures, and are highly subjective, making it difficult to achieve non-invasive, convenient, and accurate early screening and diagnosis. Furthermore, they cannot effectively differentiate between different types of muscle function decline.
A flexible dry electrode array was used to collect electromyographic signals. Combined with an adaptive notch filter algorithm and a deep learning model, multi-dimensional features were extracted to assess sarcopenia risk and determine the type of muscle function decline.
It achieves non-invasive, comfortable, and long-term stable acquisition of electromyographic signals, effectively filters out interference, improves the accuracy and specificity of diagnosis, and can accurately assess the risk of sarcopenia and the type of muscle function decline.
Smart Images
Figure CN121533739A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical testing technology, and in particular to a sarcopenic muscular dystrophy electromyography (EMG) signal acquisition and analysis system. Background Technology
[0002] Sarcopenia is a progressive, systemic syndrome characterized by a decrease in muscle mass, strength, and function associated with aging. It not only restricts mobility and increases the risk of falls in older adults but also leads to various complications, severely impacting their quality of life and increasing the healthcare burden. With the deepening of population aging, early screening and accurate diagnosis of sarcopenia have become crucial issues in the field of public health.
[0003] Currently, the diagnosis of sarcopenia mainly relies on traditional methods such as muscle mass testing (e.g., dual-energy X-ray absorptiometry, bioelectrical impedance analysis), muscle strength testing (e.g., handgrip dynamometer measurement), and walking speed assessment. However, these methods have significant limitations: muscle mass testing equipment is expensive and complex to operate, making it difficult to popularize in community healthcare or home settings; muscle strength testing and walking speed assessment are easily affected by factors such as the subject's subjective state and the testing environment, resulting in insufficient accuracy and objectivity; at the same time, traditional methods are difficult to provide early warning of sarcopenia and cannot effectively distinguish the specific types of muscle function decline, leading to a lack of targeted intervention and treatment.
[0004] Electromyography (EMG) signals, as bioelectrical signals generated by the neuromuscular system during muscle contraction, directly reflect the physiological state of muscles and neuromuscular conduction function, providing important biomarkers for the diagnosis of sarcopenia. Existing muscle function assessment technologies based on EMG signals mostly have the following defects: (1) The signal acquisition module uses rigid electrodes, which are uncomfortable to wear and are easily affected by motion artifacts, making it difficult to achieve stable acquisition over a long period of time; (2) The signal preprocessing method is singular and cannot effectively filter out various interference signals such as power frequency interference and high-frequency noise, affecting the accuracy of subsequent feature extraction; (3) The feature extraction dimensions are insufficient, mostly focusing on single-dimensional features in the time domain or frequency domain, failing to comprehensively reflect the muscle function changes related to sarcopenia; (4) The analysis and diagnostic model lacks clinical sample training and optimization, resulting in low accuracy of risk assessment and inability to accurately distinguish the types of muscle function decline. Summary of the Invention
[0005] The purpose of this invention is to provide a sarcopenia electromyography signal acquisition and analysis system to solve the technical problems of existing sarcopenia diagnostic methods, such as expensive equipment, complex operation, strong subjectivity, and inability to conduct real-time quantitative monitoring, so as to achieve non-invasive, convenient, and accurate diagnosis and risk assessment of sarcopenia.
[0006] To achieve the above objectives, the present invention provides the following solution: An electromyography (EMG) signal acquisition and analysis system for sarcopenia includes an EMG signal acquisition module, a signal preprocessing module, a feature extraction module, an analysis and diagnosis module, and a result output module, which are connected in sequence. Among them, the electromyography signal acquisition module is used to acquire the raw electromyography signals of the subject and transmit them to the subsequent processing module; The signal preprocessing module is used to filter out power frequency interference, motion artifacts and high-frequency noise in the electromyography (EMG) signal, and output the standardized effective EMG signal. The feature extraction module is used to extract multi-dimensional features related to sarcopenia from the preprocessed electromyography signals; The analysis and diagnosis module is used to assess the risk level of sarcopenia and assist in determining the type of muscle function decline based on multi-dimensional features; The results output module is used to store the detection data and generate evaluation results.
[0007] Furthermore, the electromyography signal acquisition module includes a flexible dry electrode array, an analog-to-digital converter, and a wireless transmission unit; The flexible dry electrode array is used to adhere to the surface of the target muscle group of the subject and non-invasively collect raw electromyographic signals. The analog-to-digital converter is used to convert analog electromyographic signals into digital signals; The wireless transmission unit is used to wirelessly transmit digital electromyography signals to the subsequent processing module via Bluetooth or Wi-Fi module.
[0008] Furthermore, the flexible dry electrode array is made of medical-grade conductive silicone, and the non-contact surfaces of the electrodes are provided with a shielding layer.
[0009] Furthermore, the signal preprocessing module includes a power frequency notch filter unit, a bandpass filter unit, and a signal normalization unit connected in sequence; The power frequency notch unit is used to filter out 50Hz / 60Hz power frequency interference and harmonic interference using an adaptive notch algorithm; The bandpass filter unit is used to filter out low-frequency drift and high-frequency noise; The signal standardization unit is used to eliminate individual differences and electrode placement deviations by employing the maximal-minimum normalization method, and output standardized electromyographic signals of uniform scale.
[0010] Furthermore, the multi-dimensional features include time-domain features, frequency-domain features, and nonlinear features; The time-domain features include mean absolute value, root mean square, peak factor, and kurtosis, reflecting the intensity and stability of muscle contraction; the frequency-domain features include mean power frequency, median frequency, and power spectral density, reflecting muscle fiber type and neuromuscular conduction function; the nonlinear features include approximate entropy, sample entropy, and Lempel-Ziv complexity, reflecting the irregularity and complexity of electromyographic signals, effectively distinguishing the electrical activity patterns of normal and diseased muscles.
[0011] Furthermore, the frequency domain features are extracted after the standardized electromyographic signal is converted from the time domain to the frequency domain using a fast Fourier transform.
[0012] Furthermore, the analysis and diagnosis module includes a model training unit and an assessment and diagnosis unit. The model training unit constructs a sarcopenia risk assessment model based on a deep learning model. The sarcopenia risk assessment model is trained and optimized using electromyographic signal samples from clinical sarcopenia patients and healthy individuals. The assessment and diagnosis unit outputs the sarcopenia risk level and the type of muscle function decline.
[0013] Furthermore, the sarcopenia risk level includes three levels: low, medium, and high, and the types of muscle function decline include fibromuscular atrophy, nerve innervation disorder, and metabolic abnormality.
[0014] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: This invention employs a flexible dry electrode array made of medical-grade conductive silicone material, which offers high wearing comfort and good biocompatibility. It can closely adhere to the surface of the target muscle group, enabling non-invasive and long-term stable data acquisition. The shielding layer on the non-contact surface of the electrodes and the wireless transmission design effectively reduce the impact of external electromagnetic interference and motion artifacts on signal quality, thereby improving the accuracy of the original signal acquisition.
[0015] Through the synergistic effect of the power frequency notch unit, bandpass filter unit, and signal normalization unit, the adaptive notch algorithm is used to accurately filter out 50Hz / 60Hz power frequency interference and harmonic interference, while removing low-frequency drift and high-frequency noise. Combined with the max-min normalization method, individual differences and electrode placement deviations are eliminated, and effective electromyographic signals of uniform scale are output, laying a high-quality data foundation for subsequent feature extraction.
[0016] It innovatively integrates three dimensions of features: time domain, frequency domain, and nonlinearity. The time domain features reflect the strength and stability of muscle contraction, the frequency domain features reflect the type of muscle fibers and neuromuscular transmission function, and the nonlinear features reveal the irregularity and complexity of the signal. The complementary features of the multi-dimensional features can comprehensively characterize the changes in muscle physiological state related to sarcopenia, significantly improving the specificity and sensitivity of feature recognition. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments 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 these drawings without creative effort.
[0018] Figure 1 This is a system framework diagram of the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0021] like Figure 1 As shown, the present invention provides an electromyography (EMG) signal acquisition and analysis system for sarcopenia, comprising an EMG signal acquisition module, a signal preprocessing module, a feature extraction module, an analysis and diagnosis module, and a result output module that are connected in sequence. The electromyography (EMG) signal acquisition module is used to acquire raw EMG signals from the subject and transmit them to the subsequent processing module. The EMG signal acquisition module includes a flexible dry electrode array, an analog-to-digital converter (ADC), and a wireless transmission unit. The flexible dry electrode array is used to conform to the surface of the target muscle group of the subject for non-invasive acquisition of raw EMG signals. The ADC is used to convert analog EMG signals into digital signals. The wireless transmission unit is used to wirelessly transmit digital EMG signals to the subsequent processing module via Bluetooth or Wi-Fi. The flexible dry electrode array is made of medical-grade conductive silicone, and the non-contact surfaces of the electrodes are shielded. The signal preprocessing module is used to filter out power frequency interference, motion artifacts, and high-frequency noise in the electromyography (EMG) signal, and outputs a standardized and effective EMG signal. The signal preprocessing module includes a power frequency notch filter unit, a bandpass filter unit, and a signal standardization unit connected in sequence. The power frequency notch filter unit is used to filter out 50Hz / 60Hz power frequency interference and harmonic interference using an adaptive notch filter algorithm. The bandpass filter unit is used to filter out low-frequency drift and high-frequency noise. The signal standardization unit is used to eliminate individual differences and electrode placement deviations using a maximum-minimum normalization method, and outputs a standardized EMG signal with a uniform scale. The feature extraction module is used to extract multi-dimensional features related to sarcopenia from the preprocessed electromyography (EMG) signals. These multi-dimensional features include time-domain features, frequency-domain features, and nonlinear features. The time-domain features include mean absolute value, root mean square, peak factor, and kurtosis, reflecting the intensity and stability of muscle contraction. The frequency-domain features include mean power frequency, median frequency, and power spectral density, reflecting muscle fiber type and neuromuscular conduction function. The nonlinear features include approximate entropy, sample entropy, and Lempel-Ziv complexity, reflecting the irregularity and complexity of the EMG signals and effectively distinguishing the electrical activity patterns of normal and diseased muscles. The frequency-domain features are extracted after the standardized EMG signals are converted from the time domain to the frequency domain using a fast Fourier transform. The analysis and diagnosis module is used to assess the risk level of sarcopenia and assist in determining the type of muscle function decline based on multi-dimensional features. The analysis and diagnosis module includes a model training unit and an assessment and diagnosis unit. The model training unit constructs a sarcopenia risk assessment model based on a deep learning model. The sarcopenia risk assessment model is trained and optimized using electromyographic signal samples from clinical sarcopenia patients and healthy individuals. The assessment and diagnosis unit outputs the sarcopenia risk level and the type of muscle function decline. The sarcopenia risk level includes three levels: low, medium, and high. The types of muscle function decline include fibromuscular atrophy, nerve innervation disorder, and metabolic abnormality. The results output module is used to store the detection data and generate evaluation results.
[0022] Example 1 The sarcopenia electromyography signal acquisition and analysis system of this embodiment includes an electromyography signal acquisition module, a signal preprocessing module, a feature extraction module, an analysis and diagnosis module, and a result output module that are electrically connected in sequence. The specific composition and workflow of each module are as follows: Electromyography (EMG) signal acquisition module: Utilizes a flexible dry electrode array (made of medical-grade conductive silicone with an aluminum foil shielding layer on the non-contact surface) to conform to the surface of target muscle groups such as the quadriceps and biceps of the subject, non-invasively acquiring raw analog EMG signals; The analog-to-digital conversion unit uses a 16-bit ADC chip to convert the analog signal into a digital signal with a sampling rate of 1000Hz; The wireless transmission unit transmits the digital EMG signal to the signal preprocessing module via a Bluetooth 5.0 module.
[0023] Signal preprocessing module: The power frequency notch filter unit adopts an adaptive notch filter algorithm with a center frequency set to 50Hz to filter out power frequency interference and second and third harmonic interference; the bandpass filter unit adopts a Butterworth filter with a passband range of 20-500Hz to filter out low-frequency drift (<20Hz) and high-frequency noise (>500Hz); the signal standardization unit adopts the maximum-minimum normalization method to map the filtered signal to the [0,1] interval, eliminating the influence of factors such as individual muscle mass and electrode placement deviation, and outputting a standardized electromyographic signal.
[0024] Feature extraction module: Performs multi-dimensional feature extraction on standardized electromyography signals. Temporal characteristics: Calculate the mean absolute value (MAV), root mean square (RMS), peak factor (PF), and kurtosis (K) to reflect the strength and stability of muscle contraction; Frequency domain characteristics: The time domain signal is converted to the frequency domain by Fast Fourier Transform (FFT) to calculate the average power frequency (MPF), median frequency (MF), and power spectral density (PSD), which reflect the muscle fiber type and neuromuscular conduction function. Nonlinear characteristics: Calculate the approximate entropy (ApEn), sample entropy (SampEn), and Lempel-Ziv complexity (LZC) to reflect the irregularity and complexity of electromyographic signals.
[0025] Analysis and Diagnosis Module: The model training unit constructs a sarcopenia risk assessment model based on a CNN-LSTM hybrid model. The CNN feature extraction sub-network includes two convolutional layers (layer 1 with 32 3×3 convolutional kernels and layer 2 with 64 3×3 convolutional kernels) and two max pooling layers (2×2 pooling kernels). The LSTM temporal modeling sub-network includes two bidirectional LSTM layers (layer 1 with 128 hidden units and layer 2 with 64 hidden units). The fully connected output layer includes two output branches (risk level and regression type). The model training samples include electromyographic signal data from 300 healthy individuals (control group) and 300 sarcopenia patients (100 low-risk, 100 medium-risk, and 100 high-risk). After five-fold cross-validation, the model's risk assessment accuracy is 93.2%. The assessment and diagnosis unit outputs the sarcopenia risk level and muscle function regression type (fibromuscular atrophy, nerve innervation disorder, and metabolic abnormality).
[0026] Results output module: Uses embedded storage unit to store raw signals, preprocessed signals, feature data and evaluation results; displays risk level, decline type and corresponding intervention suggestions in real time on LCD screen, and supports exporting data reports via USB interface for clinicians' reference.
[0027] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.
[0028] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.
Claims
1. A sarcopenia myoelectric signal acquisition analysis system, characterized by, The system comprises, in sequence, an electromyographic signal acquisition module, a signal preprocessing module, a feature extraction module, an analysis and diagnosis module, and a result output module. The electromyographic signal acquisition module is configured to acquire electromyographic signals of a subject and transmit the signals to the subsequent processing modules. The signal preprocessing module is configured to filter out power frequency interference, motion artifacts, and high-frequency noise in the electromyographic signals and output standardized effective electromyographic signals. The feature extraction module is configured to extract multi-dimensional features related to sarcopenia from the preprocessed electromyographic signals. The analysis and diagnosis module is configured to complete sarcopenia risk level evaluation and muscle function decline type auxiliary judgment based on the multi-dimensional features. The result output module is configured to store detection data and generate evaluation results.
2. The sarcopenic myoelectric signal acquisition and analysis system of claim 1, wherein, The electromyographic signal acquisition module comprises a flexible dry electrode array, an analog-to-digital conversion unit, and a wireless transmission unit. The flexible dry electrode array is configured to adhere to the surface of the target muscle group of the subject and non-invasively acquire electromyographic signals. The analog-to-digital conversion unit is configured to convert analog electromyographic signals into digital signals. The wireless transmission unit is configured to wirelessly transmit digital electromyographic signals to the subsequent processing modules through a Bluetooth or Wi-Fi module.
3. The sarcopenic myoelectric signal acquisition and analysis system of claim 2, wherein, The flexible dry electrode array is made of medical-grade conductive silicone, and the non-contact surface of the electrode is provided with a shielding layer.
4. The sarcopenic myoelectric signal acquisition and analysis system of claim 1, wherein, The signal preprocessing module comprises, in sequence, a power frequency notch unit, a band-pass filter unit, and a signal standardization unit. The power frequency notch unit is configured to filter out 50Hz / 60Hz power frequency interference and harmonic interference using an adaptive notch algorithm. The band-pass filter unit is configured to filter out low-frequency drift and high-frequency noise. The signal standardization unit is configured to eliminate individual differences and electrode placement deviations using a maximum-minimum normalization method and output standardized electromyographic signals of a uniform scale.
5. The sarcopenic myoelectric signal acquisition and analysis system of claim 1, wherein, The multi-dimensional features include time domain features, frequency domain features, and nonlinear features. The time domain features include average absolute value, root mean square, peak factor, and kurtosis, which reflect the intensity and stability of muscle contraction. The frequency domain features include average power frequency, median frequency, and power spectral density, which reflect muscle fiber types and neuromuscular transmission functions.
6. The sarcopenic myoelectric signal acquisition and analysis system of claim 1, wherein, The nonlinear features include approximate entropy, sample entropy, and Lempel-Ziv complexity, which reflect the irregularity and complexity of electromyographic signals and effectively distinguish the electrical activity patterns of normal muscles and diseased muscles.
7. The sarcopenic myoelectric signal acquisition and analysis system of claim 1, wherein, The frequency domain features are extracted after converting the standardized electromyographic signals from the time domain to the frequency domain through fast Fourier transform. The analysis and diagnosis module comprises a model training unit and an evaluation and diagnosis unit. The model training unit constructs a sarcopenia risk evaluation model based on a deep learning model, which is trained and optimized using electromyographic signal samples of clinical sarcopenia patients and healthy people. The evaluation and diagnosis unit outputs the sarcopenia risk level and the muscle function decline type.
8. The sarcopenia electromyographic signal acquisition and analysis system according to claim 7, wherein The sarcopenia risk level includes low, medium, and high levels, and the muscle function decline type includes muscle fiber atrophy type, nerve innervation disorder type, and metabolic abnormality type.