A noise reduction and electrode position optimization method for a single-lead electrocardiogram system

Abstract

This invention discloses a method for noise reduction and electrode placement optimization in a single-arm electrocardiogram (ECG) system, comprising the following steps: Step S1: Acquiring electromyographic (EMG) signals at multiple preset electrode positions on the upper arm of the subject, and obtaining EMG signal intensity data and EMG signal stability data for each electrode position; Step S2: Determining one or more electrode positions with lower EMG signal intensity and better signal stability from the multiple preset electrode positions, and using the one or more electrode positions as the initial electrode arrangement; Step S3: Acquiring the subject's motion state data in real time while the subject is in motion; Step S4: Dynamically selecting electrode channels with fewer EMG signals in the current motion state from the initial electrode arrangement for ECG signal acquisition. This invention can reduce the introduction of EMG signals at the source, ensuring high-quality and stable ECG signals can be obtained in different individuals and under various motion states.

Description

Technical Field

[0001] This invention relates to the field of medical and health monitoring technology, specifically to a method for noise reduction and electrode position optimization in a single-arm electrocardiogram system. Background Technology

[0002] With the increasing incidence of cardiovascular diseases, cardiac health monitoring has become an indispensable part of daily health management. Traditional electrocardiogram (ECG) devices are widely used in hospitals and clinics, collecting cardiac electrical activity signals through multiple electrodes to help doctors diagnose various heart diseases. However, traditional ECG monitoring devices are usually bulky and require multiple electrode patches, necessitating static examinations in hospital or clinical settings. While this method provides accurate ECG data, its limitations are also significant: it cannot meet the needs of long-term, daily, or home monitoring; wearing multiple electrodes and connecting to complex devices for extended periods is inconvenient and uncomfortable for patients, especially the elderly, thus affecting the continuity of wear and data accuracy; furthermore, traditional devices are difficult to use for real-time monitoring in daily life, failing to reflect the patient's cardiac health status during exercise or daily activities. Therefore, long-term and continuous ECG monitoring, especially in dynamic environments, has become a challenge for cardiac health management.

[0003] To address these issues, single-arm electrocardiography (ECG) systems were developed. A single-arm ECG is a simplified version of an ECG monitoring system, designed to provide convenient and comfortable ECG monitoring, especially during dynamic activities or at home, by reducing the number of electrodes and optimizing electrode placement. Its single-arm wearing method significantly improves the device's convenience and comfort, making it suitable for long-term, daily monitoring. However, the core challenge of single-arm ECG systems lies in their susceptibility to electromyographic interference and motion artifacts, particularly during user activity or muscle exertion, where ECG and electromyographic signals mix, severely compromising the accuracy of the ECG.

[0004] Currently, common wearable ECG monitoring devices mainly fall into the following categories: smartwatches / bracelets, which are usually equipped with heart rate sensors and monitor ECG signals in real time through built-in optical sensors or electrodes, but their accuracy is limited and cannot completely replace traditional hospital ECG equipment; chest strap ECG monitoring devices, which provide relatively accurate ECG signals by placing electrodes on the user's chest and are widely used in sports monitoring and health management, but may cause discomfort during use due to the limitation of wearing position, especially for long-term wear or daily activities.

[0005] During exercise, motion artifacts and electromyographic (EMG) noise become major problems affecting electrocardiogram (ECG) quality. Motion artifacts are usually caused by body movements such as arm swinging, walking, and running. During movement, the contact between the electrodes and the skin is unstable, leading to interference with the ECG signal. EMG noise occurs when the electrical signals generated by muscle activity overlap with the ECG signal, causing ECG signal distortion and making it difficult to accurately reflect the heart's electrical activity. In current technologies, electrode placement is mostly fixed, making it difficult to adapt to the different EMG signal characteristics and exercise states of different users. Current EMG noise signal preprocessing and data augmentation methods mainly involve preprocessing ECG signals using signal preprocessing techniques (such as detrending and denoising) or using machine learning and deep learning methods to build noise recognition and removal models to automatically identify and eliminate motion artifacts and EMG noise.

[0006] Armband-type electrocardiogram (ECG) devices, as an emerging form, offer greater comfort and adaptability compared to chest straps and watches. By placing multiple electrodes on the arm, the armband device can acquire ECG signals in real time and transmit the data to external devices, making it suitable for home monitoring scenarios requiring prolonged wear. However, electromyography (EMG) signals are currently one of the most serious problems affecting the signal quality of armband-type ECGs, and existing technologies still have significant shortcomings in EMG noise removal and electrode placement optimization.

[0007] (a) Insufficient inhibition of electromyographic noise Although existing technologies use methods such as bandpass filters and adaptive filtering to reduce the impact of electromyographic noise, the following limitations still exist: 1. Limitations of Filtering Techniques: While bandpass filters and adaptive filters can remove low-frequency EMG noise, they are less effective at removing high-frequency EMG noise. Bandpass filters cannot accurately distinguish high-frequency components in ECG and EMG signals, potentially leading to the loss of high-frequency details in the ECG and affecting accurate analysis of cardiac activity. Although adaptive filters can dynamically adjust filter parameters, their denoising effect is unstable and difficult to provide precise processing when dealing with a wide and variable frequency range of EMG noise.

[0008] 2. The Variability of Electromyographic Noise: Electromyographic noise is generated by muscle electrical activity, and its characteristics are influenced by various factors such as individual differences, exercise type, and exercise intensity, exhibiting high time-varying and individualized characteristics. Under different exercise states (such as walking, running, weightlifting, etc.), the spectrum and amplitude of electromyographic signals change significantly, making it difficult for traditional filtering techniques to effectively separate noise generated by each type of exercise or muscle activity. Existing technologies also cannot provide sufficient individualized adjustments for the electromyographic characteristics of different subjects, resulting in unstable denoising effects across different populations and difficulty adapting to electromyographic noise patterns under different exercise states.

[0009] Therefore, existing electromyographic noise removal methods suffer from insufficient accuracy, poor adaptability, and unstable denoising effects, failing to provide ideal signal quality under various complex dynamic conditions.

[0010] (ii) The optimization of electrode positions is not personalized enough. Most current wearable electrocardiogram (ECG) monitoring devices use fixed electrode placement, a design approach that has significant limitations in meeting individualized needs and accommodating dynamic movements. 1. Inability to optimize electrode placement for individuals: Fixed electrode placement does not take into account the individualized electromyographic distribution characteristics, body size differences, and individualized movement patterns of each subject. Because each person's muscle activity patterns and the location of the ECG signal source differ, a fixed electrode arrangement may lead to significant electromyographic interference in some subjects under specific movement states, affecting signal accuracy. Existing equipment cannot adjust electrode placement according to the needs of different individuals, resulting in inconsistent signal quality among different subjects.

[0011] 2. Unstable electrode contact during dynamic activities: During exercise or dynamic activities, the subject's body deforms or moves, causing the electrode contact with the skin to loosen or shift, affecting the quality of ECG signal acquisition. Especially during high-intensity exercise (such as running, weightlifting, etc.), electrodes are prone to poor contact or even detachment due to skin friction, changes in movement direction, or improper wearing. Existing electrode placement optimization is usually fixed and cannot adapt to changes caused by body movement in real time.

[0012] In summary, existing technologies primarily rely on post-processing (such as filtering and noise reduction) to remove EMG interference, but this cannot fundamentally solve the problem and often has side effects on ECG signals. Therefore, a noise reduction method is needed to reduce the introduction of EMG signals at the source. This method can effectively reduce signal errors caused by EMG interference and motion artifacts by optimizing the electrode placement design of single-arm ECG and combining it with accelerometer-based dynamic motion monitoring, ensuring that the single-arm ECG system can still provide high-quality and stable ECG signals under various motion conditions. Specifically, the following technical problems need to be addressed: how to select electrode positions with less EMG signal in the initial stage, and how to automatically switch to electrode channels with less EMG signal based on real-time motion status during exercise.

[0013] In summary, current technologies mainly rely on post-processing (such as filtering and noise reduction) to remove electromyographic interference, but this cannot fundamentally solve the problem and often has side effects on ECG signals. How to select electrode locations with less electromyographic signal in the initial stage, and how to automatically switch to electrode channels with less electromyographic signal based on real-time motion status during exercise, are unresolved issues in current single-arm ECG technology. Summary of the Invention

[0014] To overcome the shortcomings of the prior art, the present invention aims to provide a noise reduction and electrode position optimization method for a single-arm electrocardiogram system. This method can select electrode positions with less electromyographic signal in the initial stage and dynamically switch to the electrode channel with the least electromyographic interference during exercise based on the real-time exercise state and electrode signal quality. This reduces the introduction of electromyographic signal from the source and ensures that high-quality and stable electrocardiogram signals can be obtained in different individuals and under various exercise states.

[0015] To achieve the objective of this invention, the following solution is adopted: A method for noise reduction and electrode placement optimization in a single-arm electrocardiogram system includes the following steps: Step S1: Collect electromyographic signals at multiple preset electrode locations on the subject's upper arm, and obtain electromyographic signal intensity data and electromyographic signal stability data at each electrode location; Step S2: Based on the electromyographic signal intensity data and the electromyographic signal stability data, determine one or more electrode positions with low electromyographic signal intensity and good signal stability from the plurality of preset electrode positions, and use the one or more electrode positions as the initial electrode arrangement; Step S3: While the subject is in motion, collect the subject's motion state data in real time; Step S4: Based on the motion state data and the real-time signal quality acquired from each electrode in the initial electrode arrangement, dynamically select the electrode channel with less electromyographic signal in the current motion state from the initial electrode arrangement for the acquisition of electrocardiogram signal.

[0016] Furthermore, in step S4, dynamically selecting the electrode channel further includes: When the motion state data indicates a change in the subject's motion type, or when the signal quality of the currently selected electrode channel is detected to be declining, the system automatically switches to another electrode channel with more stable signal quality in the initial electrode arrangement.

[0017] Further, step S2 specifically includes: Based on the electromyographic signals collected at multiple preset electrode locations, an individualized electromyographic distribution map of the subject is generated; Based on the individualized electromyography distribution map, the electromyographic activity level and stability of each electrode location are evaluated, and the electrode location with the lowest electromyographic activity level and the highest stability is automatically selected as the initial electrode arrangement.

[0018] Furthermore, after step S2, a dynamic fine-tuning step is also included: If the signal quality of the initial electrode arrangement does not meet the preset standard during subsequent signal acquisition, the position of the initial electrode arrangement will be fine-tuned based on the real-time electromyographic signal data.

[0019] Furthermore, the real-time acquisition of the subject's motion state data specifically includes: Acceleration and motion direction data of the subject are collected in real time using accelerometers and / or gyroscopes; The acceleration and motion direction data are analyzed to identify the subject's motion type, which includes stationary, walking, running, or weightlifting.

[0020] Furthermore, in step S4, dynamically selecting the electrode channel specifically includes: Based on the identified movement type, the preset movement-electrode mapping relationship is queried, and the electrode channel that matches the current movement type and is pre-marked as having less electromyographic interference is selected; The preset motion-electrode mapping relationship is established in advance based on the analysis of the electromyographic signal intensity of each electrode channel under different motion states.

[0021] Furthermore, when collecting electromyography (EMG) and electrocardiogram (ECG) signals, flexible electrodes are used for signal acquisition. These flexible electrodes maintain dynamic and stable contact with the subject's skin through an intelligent adhesion mechanism to reduce electrode loosening or displacement caused by movement.

[0022] Furthermore, step S4 also includes a feedback optimization step: Real-time monitoring of the quality of electrocardiogram signals acquired by the selected electrode channels; When signal quality is detected to be degraded due to electromyographic interference, a switching command is generated. The switching command is used to control the electrode channel selection module to reselect the electrode channel with the least interference in the initial electrode arrangement.

[0023] Furthermore, the method also includes a personalized electrode layout step: During initial use or periodic calibration, multi-channel electromyography (EMG) signals are collected from the subject under different preset movements. Based on the multi-channel EMG signals, the individual's muscle activity patterns are analyzed, and a personalized initial electrode arrangement is automatically generated for the subject based on the muscle activity patterns to replace the default initial electrode arrangement.

[0024] Furthermore, after completing the electrode channel selection, the method also includes a signal processing step: Cardiac health analysis can be performed directly using electrocardiogram signals acquired from the selected electrode channels without the need for complex post-processing filtering or signal decomposition, thereby reducing signal processing latency.

[0025] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention reduces electromyographic interference at the source, improving signal purity. Existing technologies mainly rely on post-processing filtering or signal decomposition to remove electromyographic noise, which can easily lead to ECG signal distortion or loss of high-frequency details. This invention collects electromyographic signals from multiple preset electrode locations on the subject's upper arm in the initial stage, and actively selects electrode locations with lower electromyographic signal intensity and better stability as the initial electrode arrangement based on electromyographic signal intensity and stability data. This reduces the introduction of electromyographic signals at the signal acquisition source, avoids the side effects of post-processing on the ECG signal, and significantly improves the purity and reliability of the acquired signal.

[0026] 2. This invention enables personalized electrode placement to adapt to individual differences. Different subjects exhibit significant differences in electromyographic distribution characteristics, body type, and muscle activity patterns. This invention selects electrode positions based on the subject's own electromyographic signal characteristics, rather than using a fixed electrode arrangement. This allows for the customization of optimal initial electrode positions for each subject, effectively overcoming the problem of unstable signal quality caused by ignoring individual differences in existing technologies. It is particularly suitable for long-term home monitoring scenarios using arm-cuff single-arm electrocardiogram devices.

[0027] 3. This invention dynamically adapts to the exercise state, ensuring the quality of ECG signals during exercise. During exercise, electromyographic noise changes drastically with the type and intensity of exercise, and existing fixed electrode positions cannot cope with this change. This invention dynamically selects the electrode channel with less electromyographic signal for ECG acquisition by collecting the subject's exercise state data in real time and combining it with the signal quality obtained from each electrode in real time from the initial electrode arrangement. This allows the system to automatically switch to the optimal electrode channel between different exercise types such as walking, running, and weightlifting, effectively suppressing motion artifacts and dynamic electromyographic interference, and ensuring stable, high-quality ECG signals under various exercise states.

[0028] 4. This invention simplifies the signal processing flow and improves real-time performance. It reduces electromyographic interference at the source and dynamically maintains a low-noise channel during movement, avoiding computationally intensive operations such as complex post-processing filtering, blind source separation, or deep learning denoising. This significantly simplifies the signal processing chain, reduces system computational burden and signal output delay, facilitates real-time ECG monitoring, improves device response speed and battery life, and makes it more suitable for long-term daily wear.

[0029] 5. This invention enhances electrode contact stability and improves signal quality. The initial electrode position, selected based on signal stability data, ensures more stable contact with the skin under both static and dynamic conditions. Combined with dynamic electrode switching, this effectively reduces signal loss caused by electrode loosening or displacement, thereby improving the overall robustness of the system and user comfort. Attached Figure Description

[0030] Figure 1 This is a flowchart of the noise reduction and electrode position optimization method for the single-arm electrocardiogram system in an embodiment of the present invention; Figure 2 This is a flowchart of the initial electrode position selection method in an embodiment of the present invention; Figure 3 This is a schematic diagram of electromyographic signal intensity analysis in an embodiment of the present invention; Figure 4 This is an electromyography distribution map of the subject in an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the automatic optimization and adjustment of electrode positions in an embodiment of the present invention; Figure 6 This is a flowchart of the dynamic electrode channel selection method under motion state in an embodiment of the present invention; Figure 7 This is a schematic diagram of accelerometer motion data acquisition and motion type identification in an embodiment of the present invention; Figure 8 This is a schematic diagram of real-time switching of electrode channels in an embodiment of the present invention; Figure 9 This is a schematic diagram illustrating stable contact between the flexible electrode and the skin in an embodiment of the present invention; Figure 10 This is a schematic diagram of signal quality feedback and electrode switching adjustment in an embodiment of the present invention. Detailed Implementation

[0031] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments. It should be noted that, without conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0032] like Figure 1-10 As shown, this embodiment of the invention provides a method for noise reduction and electrode position optimization in a single-arm electrocardiogram system, comprising the following steps: Step S1: Collect electromyographic signals at multiple preset electrode locations on the subject's upper arm, and obtain electromyographic signal intensity data and electromyographic signal stability data at each electrode location.

[0033] Step S2: Based on the electromyographic signal intensity data and the electromyographic signal stability data, determine one or more electrode positions with low electromyographic signal intensity and good signal stability from the plurality of preset electrode positions, and use the one or more electrode positions as the initial electrode arrangement.

[0034] Step S3: While the subject is in motion, collect the subject's motion state data in real time.

[0035] Step S4: Based on the motion state data and the real-time signal quality acquired from each electrode in the initial electrode arrangement, dynamically select the electrode channel with less electromyographic signal in the current motion state from the initial electrode arrangement for the acquisition of electrocardiogram signal.

[0036] The noise reduction and electrode position optimization methods of the single-arm electrocardiogram system according to embodiments of the present invention will be further described in detail below.

[0037] Understandably, existing technologies primarily remove these interferences through post-processing signal processing (such as filtering and noise reduction), but this approach cannot fundamentally solve the problem and often has side effects on electrocardiogram (ECG) signals. Therefore, this invention reduces the introduction of EMG signals at the source, rather than reducing them later.

[0038] Initial electrode placement selection. In existing technologies, electrode positions are often fixed and not specifically optimized for electromyographic (EMG) noise. This leads to the introduction of strong EMG signals in many cases, affecting the quality of ECG signal acquisition. Therefore, this invention proposes a method to ensure signal purity by selecting electrode positions with less EMG signal in the initial stage. Specifically, this invention will analyze experimental data to evaluate EMG activity at different locations in the arm, selecting electrode points with smaller and more stable signals as the initial electrode positions for signal acquisition, thereby reducing EMG signal interference at the source.

[0039] Electrode channel switching during exercise. During exercise, muscle activity causes fluctuations in electromyographic (EMG) signals, making traditional electrode placement ineffective in handling dynamic changes. While existing technologies attempt to address motion artifacts through multi-channel signal fusion, the stability of electrode-skin contact during exercise cannot be guaranteed, and the fixed electrode position is difficult to adapt to changes in movement. The technical problem this invention aims to solve is how to automatically switch to electrodes with lower EMG signals based on real-time motion status, using accelerometers or other sensors to identify the type of movement. This method not only reduces EMG noise caused by movement but also flexibly selects the optimal electrode channel under different motion states, thereby ensuring stable signal quality during exercise.

[0040] This invention reduces the introduction of electromyographic (EMG) signals at the source by optimizing electrode placement and dynamically adjusting electrode selection, thereby obtaining high-quality electrocardiogram (ECG) signals under various exercise and static conditions. Specifically, it first selects electrode positions with less EMG signal in the initial stage, and then switches to electrode channels with less EMG signal in real time during exercise to reduce EMG noise interference and improve the accuracy of ECG acquisition.

[0041] In this embodiment, the selection of electrode positions in the initial stage is detailed as follows: To ensure that the single-arm electrocardiogram system can stably acquire high-quality electrocardiogram signals under both motion and static conditions, this invention designs a method for optimizing electrode placement, aiming to reduce interference from electromyographic signals at the source. Specifically, selecting electrodes with less electromyographic signal is one of the key steps of this invention.

[0042] This invention selects electrode locations with low electromyographic signals through the following steps: 1. Steps for determining the criteria for electrode placement: First, by analyzing the electromyographic (EMG) signal characteristics of the subjects, electrode placement criteria applicable to all subjects were determined. Before the experiment, multiple EMG electrodes were placed at various locations on the upper arm for short-term signal acquisition, collecting EMG signals from different locations. This process included: Electromyography (EMG) signal intensity analysis: For each electrode point, the intensity of the EMG signal was acquired and evaluated. Electrode points with lower signal intensity indicate less EMG activity and will not significantly interfere with the ECG signal. Based on this data, the system identified the electrode points with minimal EMG activity.

[0043] Electromyographic signal stability assessment: In addition to signal intensity, it is also necessary to evaluate the signal stability of electrodes at different locations. The signal stability of each electrode point is analyzed under different experimental environments (such as different motion states or resting states) to screen out electrode locations that are both stable and have low interference.

[0044] 2. Adaptation steps to address individual differences among test takers: Because electromyographic activity varies among individuals, simply relying on standard electrode placement may not meet the needs of all subjects. Therefore, this invention ensures that each subject receives optimal electrode placement before the start of the experiment through individualized electrode placement selection: Electromyography (EMG) distribution map generation: After acquiring EMG signals using electrodes at different locations using the EMG signal acquisition module, an EMG distribution map of the subject is generated. This map helps assess the EMG signal levels of the subject in various areas, thereby selecting the most suitable electrode locations.

[0045] Personalized electrode layout: Based on the subject's electromyographic distribution characteristics, the system automatically adjusts the electrode positions using algorithms, selecting the areas with the least electromyographic interference for signal acquisition.

[0046] 3. Electrode position optimization method steps: To further improve the accuracy and stability of electrode positioning, this invention optimizes electrode positioning through the following methods: Automatic Adjustment Based on Electrode Selection Criteria: This invention designs an electrode selection algorithm based on electromyographic signal intensity and stability. Once the system collects sufficient electromyographic signal data, the algorithm automatically analyzes and determines the optimal electrode location. This process is continuously optimized and corrected through experimental data feedback to ensure that electrode points with minimal electromyographic interference are selected on different subjects.

[0047] Dynamic optimization and adjustment: A dynamic adjustment mechanism is adopted during the selection of electrode positions. If it is found during the experiment that the initial electrode positions cannot effectively avoid electromyographic interference, the system can fine-tune the electrode positions based on real-time data to ensure that the selection of the initial electrode points minimizes electromyographic noise.

[0048] 4. Initial Electrode Placement Steps: Through the above steps, the optimal electrode positions were selected before the experiment began, and these positions were used as the initial electrode placement for the single-arm electrocardiogram (ECG) device. At this point, the electrode placement minimizes interference from electromyographic (EMG) signals, resulting in more stable and accurate ECG data during subsequent signal acquisition. This electrode placement is tailored to each subject's EMG characteristics and individual differences, ensuring stable contact between the electrodes and the skin and effectively avoiding areas with high EMG interference.

[0049] In this embodiment, the switching of electrode channels during the movement is detailed as follows: During exercise, electromyographic (EMG) signals generated by muscle activity can interfere with electrocardiogram (ECG) signals. Traditional static electrode placement methods often cannot effectively address the impact of EMG noise during exercise. This invention designs a dynamic electrode selection method that uses sensors such as accelerometers to identify the exercise state in real time and automatically switches to electrode channels with less EMG signal based on the type and intensity of exercise, ensuring high-quality ECG signals are acquired during exercise.

[0050] This invention selects electrodes that generate fewer electromyographic signals during movement through the following steps: 1. Real-time motion state recognition steps: First, this invention uses sensors such as accelerometers and gyroscopes to monitor the subject's motion state in real time. These sensors can capture information such as changes in the subject's acceleration, direction of motion, frequency, and intensity. By analyzing the acceleration data, the system can accurately identify the subject's type of motion, such as gait, running, or weightlifting.

[0051] Accelerometer data acquisition: The accelerometer collects the subject's acceleration data in real time, recording the intensity and state of motion. Through data processing and motion type recognition algorithms, the system can determine the subject's current motion state.

[0052] Motion state classification and recognition: Based on the acceleration characteristics of motion, the system can classify motion into different types (such as walking, running, standing still, weightlifting, etc.). The motion recognition module identifies the subject's motion state in real time and provides it to the subsequent electrode selection module.

[0053] 2. Electrode selection and switching steps: Once the motion state is identified, the electrode selection module dynamically adjusts the electrode channels for acquiring signals based on the different motion states. Specifically, the system selects electrodes with lower electromyographic signals using the following methods: Electromyography (EMG) signal analysis: Before the start of exercise, the system of this invention has already selected initial electrodes with fewer EMG signals through electrode placement optimization. Subsequently, based on real-time motion data, the system monitors the signal quality of different electrode points and analyzes the EMG signal intensity and stability of each electrode point.

[0054] Electrode adjustment based on motion state: The system automatically switches to electrodes adapted to the current motion state based on the subject's type and intensity of motion. For example, during brisk walking or running, the system selects electrodes that cause less electromyographic interference during movement; while during weightlifting or rest, the system selects the most stable electrodes with the least interference for signal acquisition.

[0055] Real-time electrode channel switching: During exercise, the switching of electrode channels is real-time. The accelerometer and signal acquisition module continuously provide feedback on motion data and electrode signal quality. The system automatically selects and switches to the electrode channel with less electromyographic noise to reduce interference caused by electromyographic activity.

[0056] 3. Electrode-skin contact stability steps: During movement, the stability of the contact between the electrode and the skin is crucial for signal quality. To ensure the stability of the electrode position during movement, this invention employs flexible materials and an intelligent adhesion mechanism in the electrode design, enabling the electrode to better adhere to the skin during movement and reducing electrode loosening or displacement caused by movement.

[0057] Flexible electrode design: This invention uses flexible electrodes that can maintain good contact with muscle movement and skin deformation, reducing poor contact and signal loss caused by movement.

[0058] Intelligent adhesion mechanism: The electrode surface coating and design ensure that it can automatically adjust according to the dynamic movement changes of the subject, optimize the contact with the skin, and improve signal stability.

[0059] 4. Feedback mechanism steps for dynamic electrode channel switching: To ensure signal accuracy during motion, the system employs a feedback mechanism when selecting electrode channels to monitor motion status and signal quality in real time. Signal quality feedback: The system dynamically adjusts the selection of electrode channels based on real-time data feedback. When a decrease in signal quality or an increase in electromyographic noise is detected in a certain electrode channel, the system will automatically switch to another electrode with more stable signal quality.

[0060] Real-time adjustment and optimization: Based on the state and duration of each exercise session, the system can continuously optimize the electrode selection strategy, improve the quality of the electrocardiogram signal, and ensure stability during long-term exercise.

[0061] The noise reduction and electrode position optimization method of the single-arm electrocardiogram system in this invention has the following advantages: 1. This invention reduces the introduction of electromyographic (EMG) signals at the source, avoiding reliance on subsequent signal processing. Existing technologies typically rely on post-processing filtering and noise reduction to remove EMG noise, but these methods cannot fundamentally reduce the introduction of EMG signals and can easily lead to ECG signal distortion. This invention optimizes electrode placement to directly reduce the introduction of EMG signals during the signal acquisition stage, selecting electrode locations with less EMG interference to ensure that the acquired ECG signals are purer and more accurate.

[0062] 2. This invention enables dynamic electrode channel selection to adapt to different exercise states. During exercise, fixed electrode positions are insufficient to effectively remove motion artifacts and electromyographic interference. This invention employs an accelerometer-based dynamic electrode channel selection method to identify the subject's exercise state in real time and automatically switch to electrode channels with lower electromyographic signals based on the type and intensity of exercise, ensuring the stability and accuracy of electrocardiogram signals during exercise.

[0063] 3. This invention achieves personalized electrode placement optimization, improving signal acquisition quality. Fixed electrode arrangements cannot be optimized based on differences in electromyography (EMG) distribution and body shape among different subjects. This invention, through personalized electrode layout, selects the most suitable electrode position based on the EMG signal distribution characteristics of each subject, reducing EMG interference at the source and ensuring that each subject can obtain high-quality electrocardiogram (ECG) signals.

[0064] 4. This invention reduces system complexity and improves real-time performance. Traditional EMG noise removal methods require complex signal processing, increasing computational burden and latency. This invention reduces EMG interference at its source, avoids complex post-processing, greatly simplifies system design, improves real-time performance and response speed, and is suitable for long-term continuous monitoring.

[0065] 5. This invention improves wearing comfort and enhances adaptability to exercise. Traditional wearable ECG devices may cause discomfort due to electrode placement and device design, especially during exercise. This invention employs a flexible electrode design and intelligent adhesion mechanism to ensure stable contact between the electrodes and the skin during exercise, reducing electrode loosening or displacement caused by exercise, significantly improving wearing comfort, and adapting to prolonged wear and various exercise conditions.

[0066] 6. This invention adapts to individual differences, providing personalized solutions. Fixed electrode placement does not fully consider individual differences, resulting in some subjects not obtaining stable signals. This invention, through personalized electrode position optimization and dynamic adjustment mechanisms, provides customized electrode placement schemes based on the electromyographic characteristics and movement patterns of different individuals, ensuring that each subject can obtain stable electrocardiogram signals under various activities and exercise states.

[0067] Experimental example: I. Experimental Objective: This experiment is used to verify the technical effectiveness of the noise reduction and electrode position optimization methods of the single-arm electrocardiogram system according to the embodiments of the present invention. The experiment will not focus on the complete experimental procedure, formula calculation methods, and detailed motion design, but will directly compare the differences between the present invention and existing technical solutions in terms of electromyographic interference suppression and single-arm electrocardiogram signal quality.

[0068] The core feature of this invention is that, before placing ECG electrodes on a single arm, electromyography (EMG) signals are collected at multiple pre-set electrode locations on the subject's upper arm. EMG signal intensity and stability data are obtained for each electrode location, generating an individualized EMG distribution map. From this map, one or more electrode locations with low EMG signal intensity and good signal stability are selected as the initial electrode placement. In this experiment, multiple pre-set electrode locations are arranged in a high-density pattern, covering approximately 20% to 80% of the anatomical region of the upper arm. Existing technologies typically use fixed, empirically-based locations or pre-set template locations for ECG electrodes, without performing the aforementioned EMG signal acquisition and individualized assessment, thus making it difficult to avoid areas with strong individualized EMG activity.

[0069] This experiment aims to demonstrate that the present invention can reduce electromyographic interference and improve the quality of single-arm ECG signals by comparing the differences between the present invention and existing technologies in terms of electromyographic interference levels near ECG electrodes, ECG signal-to-noise ratio, R-wave detection stability, proportion of effective cardiac segments, and waveform consistency with reference ECG.

[0070] II. Comparison Scheme Setup: The experiment included three comparative schemes. The invention group was the "high-density EMG screening group," where high-density EMG detection was first performed in the 20% to 80% L region of the upper arm to establish an individualized EMG interference distribution. Then, ECG electrode combinations were determined from areas with low overall EMG risk and spatial continuity. The prior art group was the "fixed experience location group," where no EMG screening was performed, and single-arm ECG electrodes were placed at preset fixed or experience locations. A high EMG control group was also included to verify that when ECG electrodes were located in areas with high EMG interference, the ECG signal quality significantly decreased.

[0071] After the electrode positions for each group were determined, single-arm ECG signals were acquired, and reference ECG signals were acquired simultaneously. Subjects completed various states including resting, natural arm swing, fist clenching, elbow flexion, elbow extension, and light-weight elbow flexion. The improvement effect of this invention compared to existing technologies was evaluated by comparing the levels of electromyographic contamination and electrocardiographic quality indicators in the ECG signals of each group.

[0072] III. Key Evaluation Indicators: This experiment mainly uses the following indicators to evaluate the differences between the present invention and the prior art: 1. Comprehensive electromyography risk score near ECG electrodes is used to evaluate the level of electromyography interference in the area where the electrodes are located. The lower the value, the more suitable the location is for placing ECG electrodes.

[0073] 2. High-frequency electromyography contamination power is used to evaluate the degree to which electromyography signals are mixed with ECG signals. The lower the value, the less the ECG signal is contaminated by electromyography.

[0074] 3. ECG signal-to-noise ratio, used to evaluate the overall quality of a single-arm ECG signal. The higher the value, the better the signal quality.

[0075] 4. R-wave detection success rate and R-wave time error are used to evaluate the stability and accuracy of heartbeat detection.

[0076] 5. Effective heartbeat segment ratio, used to evaluate the usability of ECG signals in actual monitoring.

[0077] 6. The correlation coefficient with the reference ECG is used to evaluate the consistency between the single-arm ECG waveform and the reference ECG waveform.

[0078] IV. Comparison of the present invention with existing electromyography interference techniques: In an exemplary comparative experiment, the levels of electromyographic interference near the ECG electrodes in the low EMG screening group, the fixed empirical location group, and the high EMG control group are shown in the table below. These results illustrate that the present invention can effectively identify low EMG regions before ECG electrode placement, thereby reducing EMG interference near the electrodes.

[0079] Table 1 The results above show that, compared with the prior art group, the EMG_risk near the ECG electrodes in the present invention group decreased from 0.52±0.16 to 0.29±0.09, a reduction of approximately 44.2%; the high-frequency electromyographic contamination power decreased from 0.38±0.13 to 0.21±0.08, indicating that the present invention can effectively reduce electromyographic interference near the ECG electrode location. Compared with the high EMG control group, the reduction in electromyographic interference in the present invention group is more significant, further demonstrating that the electrode placement position has a significant impact on the quality of single-arm ECG signals.

[0080] The standard calculation method for the comprehensive electromyographic risk score (EMG_risk) of candidate electromyographic detection points is explained below: EMG_risk is used to evaluate the first The risk of electromyographic interference that may be introduced when candidate locations are used as single-arm ECG electrode placement sites under various movement states. The smaller the value, the lower and more stable the electromyographic interference at the candidate location, and the more suitable it is as an ECG electrode candidate location.

[0081] 1. Symbol definition: Assume there is a total One candidate electromyography (EMG) detection point The action state. The candidate point at the th The electromyographic signals collected during each action state are recorded as follows: in, This indicates the time or sampling point number. After bandpass filtering, power frequency notch filtering, and windowing processing of the signal, the number is obtained. Electromyographic signals within a time window: in, Indicates the action state The number of time windows below. This indicates the number of sampling points within each time window.

[0082] 2. Electromyographic characteristics of a single candidate point under a single action: (1) Window root mean square value: the first The candidate point at the th The action state, the first The root mean square value of electromyography within a time window is defined as: The overall root mean square value for this action state can be taken as the average of all time windows: (2) Power in the electromyography frequency band: Let For the first Each candidate point in the action state The power spectral density of the electromyographic signal is given by the following formula: The power in the electromyographic frequency band can be defined as follows: in, Preset the electromyography frequency band, such as the 20-450Hz band commonly used in surface electromyography analysis.

[0083] (3) Peak-to-peak value: the first Each candidate point in the action state The peak-to-peak value is defined as follows: (4) Proportion of electromyographic bursts: Each candidate point in the action state The proportion of electromyographic bursts under the following conditions is defined as: in, This is an indicator function; it takes the value 1 when the condition inside the parentheses is true, and 0 otherwise. Action state The threshold for judging electromyographic bursts is set below.

[0084] (5) Stability coefficient of variation: Each candidate point in the action state The stability coefficient of variation is defined as follows: in, To prevent extremely small positive numbers with a denominator of zero. The larger the value, the more unstable the electromyographic fluctuations are at that location during that movement.

[0085] 3. Feature normalization: Since different features have different dimensions and value ranges, each feature should be normalized first. Let... Represent any feature to be normalized, for example , , , or Robust quantile normalization can be used: in, This indicates that the same subject was in an active state. The feature set of all candidate points below; and These represent the 5th percentile and the 95th percentile, respectively. This indicates that the result will be restricted to... Within the range.

[0086] After normalization, the following standardized features are obtained: 4. Single-action electromyographic interference score: No. The candidate point at the th The single-action electromyographic interference score under each action state is defined as: in, That is, the first Each candidate point in the action state The EMG_score is below; For feature weights, satisfying: In one specific implementation, the following can be adopted: At this point, substituting the above... The formula is: 5. Multi-movement comprehensive electromyography risk score (EMG_risk): No. The comprehensive electromyographic risk score of each candidate point across all movement states is defined as follows: in, Indicates the first EMG_risk of candidate points, For the first The weights of each action state satisfy: When the movement states include resting, elbow flexion, elbow extension, fist clenching, arm raising, natural arm swing, and lightly weighted elbow flexion, the corresponding single-movement scores can be recorded as follows: For everyday wear scenarios, the following comprehensive scoring method can be used: 6. Contact stability correction: If electrode-skin contact impedance is recorded simultaneously, contact impedance fluctuations can be incorporated as a penalty factor into the final risk score. Let... For the first The normalized contact instability index for each candidate point is then used to correct the overall electromyographic risk score as follows: In one specific implementation, it is possible to take ,but: in, It can be used as the basis for ranking the final candidate points. The smaller the value, the lower the overall electromyographic risk and the more stable the contact at that candidate point.

[0087] 7. Set of candidate sites for low electromyography: According to All candidate points are sorted in ascending order, and the preset proportion with the lowest score is selected as the set of candidate points with low electromyography (EMG). in, Desirable to This indicates that the candidate points with the lowest risk scores are selected from the 10% to 20% of the total.

[0088] 8. ECG electrode pair risk score: Since ECG acquisition typically requires two or more electrodes to form leads, candidate electrode pairs should be further evaluated. Let the first... The candidate point and the first If 10 candidate points constitute an ECG electrode pair, then the baseline electromyographic risk of that electrode pair is defined as: in, This is a penalty coefficient for the risk difference between two points, used to avoid situations where one electrode point has a very low electromyographic risk while the other electrode point has a high electromyographic risk.

[0089] Furthermore, considering electrode spacing, lead direction, contact impedance, and ECG detectability, a comprehensive score for the electrode pair can be defined: in, This indicates the overall score of the electrode pair; Indicates a penalty term for electrode spacing or lead direction; This indicates a penalty term for instability in electrode-pair contact resistance; Indicates ECG detectability penalty; , and These are the weights of the corresponding penalty items.

[0090] Ultimately, the electrode spacing, lead direction, contact stability, and ECG detectability requirements were selected, and The lower electrode pair is used as the single-arm ECG acquisition electrode combination.

[0091] 9. Calculation Example: Suppose that the normalized feature of a candidate point under natural arm swing motion is: The electromyographic interference score of this candidate point under natural arm swing is: If the normalized feature of another candidate point under the same action is: The single-action electromyographic interference score is then: Therefore, it can be seen that the first candidate point has a lower electromyographic risk under natural arm swing, while the second candidate point has a higher electromyographic risk.

[0092] V. Comparison of ECG signal quality between the present invention and existing technologies: Further comparison of ECG signal quality under different electrode arrangements yields the expected results shown in the table below.

[0093] Table 2 As can be seen from the above results, compared with the prior art group, the ECGSNR of the present invention group increased from 10.6±3.5dB to 14.8±3.1dB; the R wave detection success rate increased from 91.2%±6.1% to 97.6%±2.3%; the R wave time error decreased from 15.6±7.8ms to 8.7±4.1ms; the proportion of effective cardiac beat segments increased from 84.9%±8.3% to 94.3%±4.5%; and the correlation coefficient with the reference ECG increased from 0.72±0.12 to 0.86±0.07.

[0094] The above results demonstrate that by performing high-density EMG screening before electrode placement, this invention can effectively reduce electromyographic contamination in single-arm ECG signals, and improve ECG waveform quality, heartbeat detection stability, and signal availability.

[0095] VI. Comparison results under dynamic motion states: To verify the effectiveness of this invention in dynamic wearing scenarios, the proportion of effective cardiac segments under various action states was further compared. The expected results are shown in the table below.

[0096] Table 3 The results above show that, under resting conditions, both the present invention group and the prior art group achieved a high proportion of effective cardiac fragments, but the present invention group still outperformed the prior art group. Under dynamic conditions characterized by enhanced electromyographic activity, such as natural arm swing, fist clenching, elbow flexion, elbow extension, and light-weight elbow flexion, the advantage of the present invention group was even more pronounced. Especially under light-weight elbow flexion, the proportion of effective cardiac fragments in the present invention group was 88.9%, significantly higher than the 72.4% in the prior art group and the 49.7% in the high EMG control group.

[0097] The results show that when ECG electrodes are placed in fixed empirical positions in the prior art, they are easily affected by individual differences in muscle activity and changes in movement state; while the present invention can maintain higher ECG signal availability in dynamic wearing scenarios by pre-screening low electromyographic regions.

[0098] VII. Conclusions of the Comparative Experiment: The comparative experimental results above show that, compared with the existing technology that does not perform electromyography screening and simply arranges ECG electrodes according to fixed empirical positions, the present invention can significantly reduce the level of electromyography interference near the ECG electrodes, reduce the mixing of high-frequency electromyography components into the ECG signal, and improve the ECG signal-to-noise ratio, R-wave detection success rate, effective cardiac segment ratio, and waveform consistency with the reference ECG.

[0099] The main technical advantages of this invention are: obtaining individualized electromyographic interference distribution in subjects through high-density EMG screening, and determining ECG electrode combinations from continuous low electromyographic regions, thereby avoiding areas with strong electromyographic activity at the ECG signal acquisition source. Unlike existing technologies that rely on filtering or post-processing to reduce interference after acquisition at a fixed location, this invention can reduce electromyographic interference entering the ECG acquisition link during the electrode placement stage, improving the stability of single-arm ECG in dynamic wearing and long-term monitoring scenarios.

[0100] Therefore, this experiment can be used to support the beneficial effects of the present invention compared to the prior art: that is, the present invention can achieve individualized optimization of the position of single-arm ECG electrodes through high-density electromyography screening, and obtain ECG acquisition quality that is superior to fixed empirical electrode placement schemes.

[0101] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A noise reduction and electrode position optimization method for a single- arm electrocardiogram system, characterized in that, Includes the following steps: Step S1: Collect electromyographic signals at multiple preset electrode locations on the subject's upper arm, and obtain electromyographic signal intensity data and electromyographic signal stability data at each electrode location; Step S2: Based on the electromyographic signal intensity data and the electromyographic signal stability data, determine one or more electrode positions with low electromyographic signal intensity and good signal stability from the plurality of preset electrode positions, and use the one or more electrode positions as the initial electrode arrangement; Step S3: While the subject is in motion, collect the subject's motion state data in real time; Step S4: Based on the motion state data and the real-time signal quality acquired from each electrode in the initial electrode arrangement, dynamically select the electrode channel with less electromyographic signal in the current motion state from the initial electrode arrangement for the acquisition of electrocardiogram signal.

2. The method of claim 1, wherein, In step S4, dynamically selecting the electrode channel further includes: When the motion state data indicates a change in the subject's motion type, or when the signal quality of the currently selected electrode channel is detected to be declining, the system automatically switches to another electrode channel with more stable signal quality in the initial electrode arrangement.

3. The method of claim 1, wherein, Step S2 specifically includes: Based on the electromyographic signals collected at multiple preset electrode locations, an individualized electromyographic distribution map of the subject is generated; Based on the individualized electromyography distribution map, the electromyographic activity level and stability of each electrode location are evaluated, and the electrode location with the lowest electromyographic activity level and the highest stability is automatically selected as the initial electrode arrangement.

4. The method of claim 1, wherein, Following step S2, a dynamic fine-tuning step is also included: If the signal quality of the initial electrode arrangement does not meet the preset standard during subsequent signal acquisition, the position of the initial electrode arrangement will be fine-tuned based on the real-time electromyographic signal data.

5. The method of claim 1, wherein, The real-time acquisition of the subject's motion state data specifically includes: Acceleration and motion direction data of the subject are collected in real time using accelerometers and / or gyroscopes; The acceleration and motion direction data are analyzed to identify the subject's motion type, which includes stationary, walking, running, or weightlifting.

6. The method of noise reduction and electrode position optimization for a single- arm electrocardiogram system of claim 5, wherein, In step S4, dynamically selecting the electrode channel specifically includes: Based on the identified movement type, the preset movement-electrode mapping relationship is queried, and the electrode channel that matches the current movement type and is pre-marked as having less electromyographic interference is selected; The preset motion-electrode mapping relationship is established in advance based on the analysis of the electromyographic signal intensity of each electrode channel under different motion states.

7. The method of noise reduction and electrode position optimization for a single- arm ECG system of claim 1, wherein, When collecting electromyography (EMG) and electrocardiogram (ECG) signals, flexible electrodes are used for signal acquisition. The flexible electrodes maintain dynamic and stable contact with the subject's skin through an intelligent adhesion mechanism to reduce electrode loosening or displacement caused by movement.

8. The method of noise reduction and electrode position optimization for a single- arm ECG system of claim 1, wherein, Step S4 also includes a feedback optimization step: Real-time monitoring of the quality of electrocardiogram signals acquired by the selected electrode channels; When signal quality is detected to be degraded due to electromyographic interference, a switching command is generated. The switching command is used to control the electrode channel selection module to reselect the electrode channel with the least interference in the initial electrode arrangement.

9. The method of noise reduction and electrode position optimization for a single- arm ECG system of claim 1, wherein, The method also includes a personalized electrode layout step: During initial use or periodic calibration, multi-channel electromyography (EMG) signals are collected from the subject under different preset movements. Based on the multi-channel EMG signals, the individual's muscle activity patterns are analyzed, and a personalized initial electrode arrangement is automatically generated for the subject based on the muscle activity patterns to replace the default initial electrode arrangement.

10. The method of claim 1, wherein, After selecting the electrode channel, the method also includes a signal processing step: Cardiac health analysis can be performed directly using electrocardiogram signals acquired from the selected electrode channels without the need for complex post-processing filtering or signal decomposition, thereby reducing signal processing latency.

Images