Wearable underwear heart rate detection method and system

Abstract

The invention relates to the technical field of heart rate detection, in particular to a wearable underwear heart rate detection method and system.The method comprises the following steps that motion information of a user is obtained, and specific body posture changes are recognized based on the motion information of the user; generating a physiological artifact signal related to the specific body posture change based on the specific body posture change; removing a physiological artifact signal from the original physiological signal to obtain a purified physiological signal; heart rate abnormity judgment is conducted on the purified physiological signals. According to the scheme, the physiological artifact signals can be effectively removed, and the accuracy of heart rate detection is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of heart rate detection, and in particular to a method and system for detecting heart rate while wearing underwear. Background Art

[0002] Currently, one of the core functions of wearable health monitoring devices is to analyze heart rate data in real time to identify potential anomalies. However, as users engage in daily activities, small, irregular movements of the device can cause changes in the contact state between the sensor and the skin, introducing non-periodic motion artifacts. Summary of the Invention

[0003] The purpose of the present invention is to address the above-mentioned shortcomings and provide a method and system for detecting heart rate while wearing underwear.

[0004] The present invention adopts the following technical solutions:

[0005] A method for detecting heart rate while wearing underwear, the method comprising the following steps:

[0006] Acquiring the user's motion information and identifying specific body posture changes based on the user's motion information;

[0007] generating a physiological artifact signal related to the specific body posture change based on the specific body posture change;

[0008] Removing physiological artifact signals from original physiological signals to obtain purified physiological signals;

[0009] Determine abnormal heart rate based on purified physiological signals.

[0010] Through the above solution, physiological artifact signals can be effectively removed, the accuracy of heart rate detection can be improved, and omissions caused by data discarding can be avoided.

[0011] Optionally, the present application further proposes that, based on a specific body posture change, the step of generating a physiological artifact signal related to the specific body posture change includes:

[0012] Extracting dynamic parameters related to specific body posture changes from the user's motion information;

[0013] Based on specific body posture changes and corresponding dynamic parameters, a physiological artifact signal is generated through an artifact prediction model, wherein at least one of the morphology, amplitude and duration of the physiological artifact signal is dynamically adjusted according to the dynamic parameters.

[0014] Through the above scheme, the shape, amplitude and duration of the physiological artifact signal can be dynamically adjusted according to the dynamic parameters, making the artifact removal more accurate.

[0015] Optionally, the present application further proposes that after the step of removing physiological artifact signals from the original physiological signals to obtain purified physiological signals, the method further includes:

[0016] Identify and purify the continuous fluctuation signal component in the physiological signal, where the continuous fluctuation signal component is a non-pulsatile signal, and the frequency of the continuous fluctuation signal component partially overlaps with the heart rate frequency range;

[0017] Suppression processing is performed on continuously fluctuating signal components.

[0018] Through the above scheme, the continuous fluctuation signal components that partially overlap with the heart rate frequency range can be further identified and suppressed, further improving the signal purity.

[0019] Optionally, the present application further proposes that the step of determining abnormal heart rate based on the purified physiological signal includes:

[0020] Obtain and update the user's heart rate range and heart rate variability parameters under different activity states to form a parameter set;

[0021] Identify the user's current activity status;

[0022] According to the user's current activity state, a heart rate range and heart rate variability parameter corresponding to the user's current activity state are selected from the parameter set as a criterion for determining heart rate abnormality;

[0023] Identify heart rate beats from purified physiological signals;

[0024] Compare heart rate pulses with the criteria for judging abnormal heart rate;

[0025] Based on the comparison results, it is determined whether there is an abnormal heart rate.

[0026] Through the above solution, the heart rate abnormality judgment standard can be dynamically adjusted according to the user's current activity status, thereby improving the accuracy and adaptability of abnormality judgment.

[0027] Optionally, the present application further proposes obtaining and updating the heart rate range and heart rate variability parameters of the user in different activity states, and the steps of forming a parameter set include:

[0028] Acquire the user's original mechanical shock signal, pre-process the original mechanical shock signal, and obtain a pre-processed mechanical shock signal;

[0029] Identify heart rate beats in purified physiological signals;

[0030] Taking the time point of occurrence of the heart beat as the starting point, within a preset time window after the starting point, characteristic events synchronized with the heart beat are identified from the pre-processed mechanical shock signal;

[0031] Calculating the time difference between the occurrence time of the heart beat and the occurrence time of the characteristic event;

[0032] Determine whether the time difference falls within the time threshold to confirm the authenticity of the heart rate beat;

[0033] The confirmed authenticity of the heart rate pulse is used to obtain and update the user's heart rate range and heart rate variability parameters under different activity states to form a parameter set;

[0034] The parameter set is limited to data corresponding to the case where the purified physiological signal is not determined to be an abnormal heart rate.

[0035] Through the above solution, the authenticity of the heart rate pulsation can be confirmed through the characteristic event of synchronization between the mechanical impact signal and the heart rate pulsation, thereby ensuring the reliability of the parameter set used for judging abnormal heart rate.

[0036] Optionally, the present application further proposes that the step of performing suppression processing on the continuous fluctuation signal component includes:

[0037] Perform frequency analysis on the purified physiological signals to extract the dominant frequency characteristics and energy distribution range of the continuous fluctuation signal components;

[0038] Configure the filter parameters based on the dominant frequency characteristics and energy distribution range of the continuously fluctuating signal components;

[0039] The configured filter is applied to purify the physiological signal to suppress the continuously fluctuating signal components.

[0040] Through the above scheme, the filter parameters can be configured based on the dominant frequency characteristics of the continuous fluctuation signal component and its energy distribution range, so as to achieve accurate suppression of the continuous fluctuation signal component.

[0041] Optionally, the present application also proposes that, in the process of suppressing the continuous fluctuation signal component, the integrity of the pulsation waveform in the purified physiological signal is monitored, and when the pulsation waveform distortion is detected, the suppression strength of the filter is reduced.

[0042] Through the above scheme, the integrity of the beat waveform can be monitored during the suppression process to avoid distortion of the real heart rate signal caused by excessive filtering.

[0043] Optionally, the present application further proposes that the step of monitoring the integrity of the pulsating waveform in the purified physiological signal includes:

[0044] Identify heart rate beats from purified physiological signals;

[0045] Extracting pulse characteristic parameters from the heart rate pulse, the pulse characteristic parameters including peak amplitude, waveform width and rising edge slope;

[0046] Get the user's current wearing status information;

[0047] Dynamically adjust the reference range of pulsation characteristic parameters based on the user's current wearing status information;

[0048] comparing the beat characteristic parameter with a reference range of the adjusted beat characteristic parameter;

[0049] Based on the comparison result, the integrity of the beat waveform in the purified physiological signal is determined.

[0050] Through the above scheme, it is possible to accurately determine the integrity of the pulsation waveform by extracting pulsation characteristic parameters and dynamically adjusting the reference range according to the wearing status.

[0051] Optionally, the present application further proposes that the step of dynamically adjusting the reference range of the pulsation characteristic parameter according to the user's current wearing state information includes:

[0052] Collect inertial measurement data and strain data at multiple locations in the underwear;

[0053] Determine the user's current wearing status based on a preset wearing status recognition model;

[0054] The corresponding wearing status label is output according to the user's current wearing status, and the wearing status label is used for adjusting the reference range of the pulsation characteristic parameters.

[0055] Through the above scheme, the user's wearing status can be accurately judged by collecting inertial measurement data and strain data and based on the wearing status recognition model, providing a basis for the dynamic adjustment of the pulsation characteristic parameters.

[0056] Optionally, the present application further proposes a heart rate detection system for wearing underwear, which is applied to the above-mentioned heart rate detection method for wearing underwear, and the system comprises:

[0057] an information processing module, configured to obtain the user's motion information and identify specific body posture changes based on the user's motion information;

[0058] A physiological artifact signal generation module, which generates a physiological artifact signal related to a specific body posture change based on the specific body posture change;

[0059] A physiological signal purification module is used to remove physiological artifact signals from original physiological signals to obtain purified physiological signals;

[0060] The heart rate abnormality judgment module judges the heart rate abnormality of the purified physiological signal.

[0061] Through the above solution, a system for implementing the above heart rate detection method is provided, which is convenient for practical application and deployment.

[0062] From the above, it can be seen that the present application provides a wearable underwear heart rate detection method and system, which obtains motion information to identify changes in body posture, generates and removes physiological artifact signals based on this, and then judges heart rate abnormalities on the purified signals, effectively solving the technical problem of accurately identifying heart rate abnormalities and avoiding missed reports in complex data streams. It has the advantages of being able to effectively remove physiological artifact signals, improve the accuracy of heart rate detection, and avoid missed reports caused by data discard.

[0063] To further understand the features and technical contents of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are only for reference and illustration and are not intended to limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a flow chart of a method for detecting heart rate while wearing underwear according to the present invention;

[0065] Figure 2 The figure is a structural diagram of a wearable underwear heart rate detection system of the present invention. DETAILED DESCRIPTION

[0066] The following is an explanation of the embodiments of the present invention through specific embodiments. Those skilled in the art can understand the advantages and effects of the present invention from the contents disclosed in this specification. The present invention can be implemented or applied through other different specific embodiments, and the details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of the present invention. In addition, the drawings of the present invention are only for simple schematic illustrations and are not depicted according to actual dimensions. It is stated in advance. The following embodiments will further explain the relevant technical contents of the present invention in detail, but the disclosed contents are not intended to limit the scope of protection of the present invention.

[0067] This embodiment provides a method and system for detecting heart rate while wearing underwear. Figure 1 and Figure 2 shown.

[0068] refer to Figure 1 A method for detecting heart rate while wearing underwear, comprising the following steps:

[0069] Acquiring the user's motion information and identifying specific body posture changes based on the user's motion information;

[0070] generating a physiological artifact signal related to the specific body posture change based on the specific body posture change;

[0071] Removing physiological artifact signals from original physiological signals to obtain purified physiological signals;

[0072] Determine abnormal heart rate based on purified physiological signals.

[0073] Motion information refers to data reflecting the user's physical activity status. It can be implemented using inertial measurement unit data, accelerometer data, gyroscope data, or pressure sensor data, such as data collected through sensors built into wearable underwear. Its main purpose is to capture the user's body behavior and provide data for identifying posture changes. Specific body posture changes refer to body posture or movement transitions associated with physiological artifacts, either predefined or identified through learning models. This can be implemented using threshold-based accelerometer signal analysis, pattern recognition algorithms, or machine learning models. Examples include transitions from sitting to standing, arm raising, or torso bending. Its main purpose is to identify behavioral patterns that may cause physiological artifacts. Physiological artifact signals refer to interference components superimposed on the original physiological signals caused by non-cardiac physiological activities or body movements. This can be implemented using artifact prediction models, signal synthesis techniques, or artifact template libraries based on empirical data. Examples include signal fluctuations caused by muscle contraction, changes in vascular pressure, or poor contact between the sensor and the skin. Its main purpose is to simulate and characterize the interference caused by specific body posture changes for subsequent removal. Raw physiological signals refer to unprocessed bioelectric or optical signals collected directly from the user's body, containing heart rate information and various interferences. These signals can be acquired using photoplethysmography (PPG) sensors, electrocardiogram (ECG) sensors, or bioelectrical impedance sensors. For example, optical signals collected by PPG sensors integrated into wearable underwear are used to provide a data stream for heart rate detection. Purified physiological signals refer to physiological data that has been processed to remove or suppress physiological artifacts. These signals can be achieved using techniques such as adaptive filtering, signal decomposition and reconstruction, or artifact subtraction. For example, a signal is obtained by subtracting predicted artifact signals from the raw PPG signal. This is primarily intended to provide a signal that is closer to the heart rate pulsation, improving the accuracy of subsequent heart rate analysis. Heart rate anomaly determination refers to the process of evaluating whether the heart rate pulsation identified in the purified physiological signals deviates from the normal physiological range or pattern based on set standards or models. This can be achieved using methods based on heart rate range comparison, heart rate variability analysis, or machine learning classifiers. For example, the current heart rate is compared with the user's historical heart rate data or medical standards. This is primarily intended to identify potential cardiovascular health issues in the user.

[0074] The core innovation of this application lies in that by combining the acquisition of user motion information and the identification of specific body posture changes with the generation of physiological artifact signals based on specific body posture changes for prediction and removal, the problem that existing methods are difficult to distinguish between heart rate pulsations and physiological artifacts when the user's body posture changes is solved, thereby achieving the effect of improving the accuracy of heart rate detection in sports scenarios and avoiding the underreporting of heart rate abnormalities due to data discarding.

[0075] Specifically, the solution of this application obtains user motion information, such as data from an inertial measurement unit, to identify the user's current specific body posture changes. This identification process is useful because it enables the system to address physiological interference caused by different posture changes. Based on the identified specific body posture changes, the system can generate physiological artifact signals related to the posture changes. This generation process does not simply apply a general model; instead, it dynamically simulates the artifact's shape, amplitude, and duration based on the posture changes, thereby reflecting the actual interference. The generated physiological artifact signals are then removed from the original physiological signals, for example, through signal subtraction or adaptive filtering techniques, to obtain a purified physiological signal. This step eliminates the interference introduced by body posture changes, making the signal closer to the physiological pulse. Finally, the purified physiological signal is used to determine heart rate anomalies. Because the signal has been processed for artifact removal, the judgment process can avoid misidentifying artifacts as heart rate anomalies, thereby improving the accuracy of heart rate anomaly detection. It is precisely because of this strategy of predicting and removing artifacts based on posture changes that the system can still stably acquire and analyze heart rate data while the user is engaged in daily activities, especially when the body posture changes, avoiding the problem of abnormal omissions caused by having to discard data due to data interference in existing methods.

[0076] In some embodiments, the present application is implemented as follows: First, an inertial measurement unit integrated into the wearable underwear continuously collects acceleration and angular velocity data from the user, which constitutes the user's motion information. A built-in posture recognition algorithm, such as a model based on a support vector machine or decision tree, analyzes this motion information in real time to identify specific body posture changes, such as a transition from sitting to standing, an arm raise, or a torso bend. Once a specific body posture change is identified, such as a user switching from sitting to standing, the system generates a physiological artifact signal associated with this posture change based on an artifact prediction model. This model can be a signal generator that synthesizes a fluctuating signal with a specific frequency and amplitude, such as a fluctuating signal, based on dynamic parameters such as the amplitude and duration of the posture change. Subsequently, a filtering algorithm or signal subtraction is used to remove the previously generated physiological artifact signal from the raw physiological signal collected by the photoplethysmography (PPG) sensor in the wearable underwear, thereby obtaining a purified physiological signal. For example, a least mean square (LMS) filter can use the generated artifact signal as input to separate the artifact component from the raw PPG signal. Finally, the purified physiological signals are analyzed for heart rate anomalies. This is achieved by identifying waveforms in the purified signals, calculating the heart rate, and comparing it to the user's heart rate range. For example, if the heart rate exceeds the user's current activity range for several consecutive seconds, the system will determine that there is a heart rate anomaly.

[0077] This application further proposes that the steps of generating a physiological artifact signal related to a specific body posture change based on the specific body posture change include:

[0078] Extracting dynamic parameters related to specific body posture changes from the user's motion information;

[0079] Based on specific body posture changes and corresponding dynamic parameters, a physiological artifact signal is generated through an artifact prediction model, wherein at least one of the morphology, amplitude and duration of the physiological artifact signal is dynamically adjusted according to the dynamic parameters.

[0080] Among them, motion information refers to data that can reflect the user's body motion state, which can be achieved by using data collected by inertial measurement unit (IMU) sensors such as accelerometers, gyroscopes, magnetometers, or data obtained through other means such as visual sensors and pressure sensors. Its purpose is to provide basic data for identifying changes in body posture and extracting dynamic parameters; specific body posture change refers to the process of the user's body changing from one posture to another, such as from sitting to standing, from still to walking, or raising the arm, etc. It can use a preset posture recognition algorithm and combine it with the characteristic patterns in the motion information for identification. Its purpose is to determine the specific scenario where a physiological artifact signal needs to be generated; dynamic parameters refer to the numerical values ​​that can quantify the kinematic characteristics such as the intensity, speed, acceleration or duration of the change in a specific body posture process. It can use the amplitude, rate of change, frequency component or energy distribution of the motion information. The purpose is to provide refined input for the artifact prediction model to achieve dynamic adjustment of the artifact signal; the artifact prediction model refers to a computational model that can predict the characteristics of the physiological artifact signal based on the input data, which can be implemented by a regression model based on machine learning, a neural network model (such as a recurrent neural network or a long short-term memory network), or a lookup table or mathematical function established based on empirical data, and its purpose is to generate corresponding physiological artifact signals according to changes in body posture and its dynamic parameters; the morphology, amplitude and duration of the physiological artifact signal refer to the waveform characteristics, signal strength and duration of the physiological artifact signal in the time domain, which can be described by the waveform shape, peak size, root mean square value and time span of the signal from the beginning to the end, and its purpose is to ensure that the generated artifact signal can more accurately simulate the characteristics of the real physiological artifact, thereby improving the accuracy of artifact removal.

[0081] The solution of this application improves the accuracy of heart rate detection by finely controlling the generation process of physiological artifact signals. Specifically, after identifying a specific body posture change, the system further extracts dynamic parameters related to that specific body posture change from the user's motion information. These dynamic parameters, such as the intensity, speed, or duration of the movement, can quantify the specific characteristics of the body posture change. The system then generates a physiological artifact signal using an artifact prediction model based on the identified specific body posture change and these extracted dynamic parameters. This artifact prediction model does not simply generate an artifact signal with a fixed pattern; instead, it dynamically adjusts the shape, amplitude, or duration of the generated physiological artifact signal based on the input dynamic parameters. For example, when the dynamic parameters indicate a more drastic body posture change, the artifact prediction model can generate a physiological artifact signal with a larger amplitude and longer duration; conversely, when the dynamic parameters indicate a more gradual body posture change, it generates an artifact signal with a smaller amplitude and shorter duration. This dynamic adjustment mechanism enables the generated physiological artifact signal to more accurately simulate the physiological artifacts caused by varying degrees of body posture change in real-world scenarios. This precise way of generating physiological artifact signals is closely combined with the step of removing physiological artifact signals from the original physiological signals in this application, forming a more complete artifact removal process. By providing an artifact signal that is highly matched with the actual physiological artifact, the artifact component can be more effectively subtracted or suppressed from the original physiological signal, thereby obtaining a purer physiological signal. This pure physiological signal provides a high-quality data basis for subsequent abnormal heart rate judgment, significantly reducing the risk of misjudgment or omission due to residual artifacts, thereby improving the overall reliability and accuracy of heart rate detection. It is precisely because of this improvement in the accuracy of artifact signal generation that the entire heart rate detection method can still maintain a high performance in complex motion scenarios, effectively solving the problem of incomplete artifact removal in traditional methods leading to a decrease in heart rate detection accuracy.

[0082] In some preferred embodiments, the present application is implemented as follows: When an inertial measurement unit (IMU) sensor, such as a three-axis accelerometer and a three-axis gyroscope, in a wearable underwear worn by a user collects the user's motion information, the system can first pre-process this motion information to identify specific body posture changes, such as the user switching from a sitting position to a standing position. Once this specific body posture change is identified, the system further extracts dynamic parameters related to this posture change from the accelerometer data. For example, the peak acceleration during the posture change, the root mean square value of the acceleration, or the duration of the posture change can be calculated. Subsequently, the system can use a pre-trained artifact prediction model to generate a physiological artifact signal. The artifact prediction model can be a recurrent neural network (RNN) based on deep learning, such as a long short-term memory network (LSTM), whose input includes the type of specific body posture change identified and the extracted dynamic parameters. The model is trained to learn the complex mapping relationship between different body posture changes and their dynamic parameters and the corresponding physiological artifact signals. When generating physiological artifact signals, the model can dynamically adjust the shape, amplitude, or duration of the generated artifact signals based on the input dynamic parameters. For example, if the extracted acceleration peak is high, indicating a more drastic posture change, the artifact prediction model can generate a physiological artifact signal with a larger amplitude and a steeper waveform; if the posture change lasts for a long time, the model can extend the duration of the generated artifact signal accordingly. In this way, the generated physiological artifact signal can more accurately reflect the physiological artifacts caused by the user's actual movement intensity and duration, thereby providing a highly matched reference signal for the subsequent artifact removal step.

[0083] The present application further proposes removing physiological artifact signals from original physiological signals to obtain purified physiological signals, further comprising:

[0084] Identify and purify the continuous fluctuation signal component in the physiological signal, where the continuous fluctuation signal component is a non-pulsatile signal, and the frequency of the continuous fluctuation signal component partially overlaps with the heart rate frequency range;

[0085] Suppression processing is performed on continuously fluctuating signal components.

[0086] Identifying the sustained fluctuation signal component in the purified physiological signal refers to analyzing the purified physiological signal to distinguish non-pulsatile interference signals whose frequencies partially overlap with the heart rate frequency range. Specifically, this can be done through time domain or frequency domain analysis methods. For example, wavelet transform, Fourier transform, or adaptive filtering techniques can be used to detect whether the signal contains sustained non-pulsatile fluctuations within a specific frequency range. The purpose is to accurately locate and distinguish the specific type of noise that interferes with heart rate detection. Sustained fluctuation signal components are non-pulsatile signals, and their frequencies partially overlap with the heart rate frequency range. Specifically, these signals are not generated by the periodic beating of the heart, but rather by other physiological activities or external interference, such as muscle tremors, respiratory movements, or environmental noise. Their frequency characteristics overlap with the frequency range of the normal heart rate, making it difficult for traditional heart rate filtering methods to completely separate them. The purpose is to clearly identify the characteristics of the interference signal that needs to be processed so that targeted suppression strategies can be adopted. Suppressing continuously fluctuating signal components involves applying appropriate signal processing techniques to identify these components to reduce or eliminate their impact on the purified physiological signal. This can be achieved through methods such as digital filtering, adaptive noise cancellation, or signal reconstruction. The goal is to minimize interference from these components without compromising the true heart rate signal, thereby improving the purity of the heart rate signal.

[0087] The solution of this application ensures the accuracy of heart rate detection by further refining the residual interference in the signal based on the initial purification of the physiological signal. Specifically, after removing the physiological artifact signals caused by specific body posture changes from the original physiological signal and obtaining the initial purified physiological signal, the solution does not stop there. Instead, it further analyzes the purified physiological signal in depth to identify possible continuous fluctuating signal components. These components are clearly defined as non-pulsatile signals, and their frequencies partially overlap with the heart rate frequency range. This means that they do not originate from cardiac pulsation, but are easily confused with the true heart rate signal, causing significant interference to subsequent heart rate detection. It is precisely because of the special nature of this signal component, namely its non-pulsatile nature and overlap with the heart rate frequency, that it is difficult to completely eliminate it through conventional artifact removal methods. Therefore, after identifying these specific continuous fluctuating signal components, this solution performs targeted suppression processing on them. This suppression processing aims to minimize the amplitude of these interfering signals while preserving the true heart rate pulsation information as much as possible. Through this two-stage signal purification strategy—first removing physiological artifacts related to body posture, then identifying and suppressing other continuously fluctuating signal components—this solution can obtain a purer physiological signal. This enables the subsequent step of determining heart rate anomalies in the purified physiological signal to be based on higher-quality data, significantly reducing the risk of misjudgments or missed detections, thereby improving the reliability and accuracy of heart rate detection. This layered, refined signal processing flow effectively addresses the problem of traditional methods struggling to completely eliminate all interference in complex noisy environments, resulting in limited heart rate detection accuracy.

[0088] In some preferred embodiments, after removing physiological artifacts from the original physiological signal and obtaining a purified physiological signal, the following steps can be specifically implemented. First, to identify persistent fluctuating signal components in the purified physiological signal, frequency analysis can be performed on the purified physiological signal. For example, a fast Fourier transform or short-time Fourier transform can be used to obtain a signal spectrogram, thereby observing the signal's energy distribution at different frequencies. By analyzing the spectrogram, persistent fluctuations that have significant energy peaks but are not caused by pulsation within the heart rate frequency range (for example, the resting heart rate of an adult is typically between 60 and 100 beats per minute, corresponding to 1 Hz to 1.67 Hz) can be identified. These fluctuations may manifest as specific narrowband or broadband noise, and their dominant frequency characteristics and energy distribution range can be extracted. Furthermore, to suppress these identified persistent fluctuating signal components, filter parameters can be configured based on their dominant frequency characteristics and energy distribution range. For example, if a persistent trembling noise of a specific frequency is identified, an adaptive filter can be configured with its center frequency set to the dominant frequency of the noise and its bandwidth adjusted based on its energy distribution range. Alternatively, if the noise manifests itself over a wider frequency range, a band-stop filter or Kalman filter can be used. This configured filter can be applied to purify the physiological signal, effectively attenuating or eliminating continuously fluctuating signal components while minimizing the integrity of the heart rate signal. In practical applications, for example, a digital signal processor or microcontroller can be used to perform these frequency analysis and filtering operations, enabling real-time or near-real-time signal processing.

[0089] This application further proposes that the steps of determining abnormal heart rate based on purified physiological signals include:

[0090] Obtain and update the user's heart rate range and heart rate variability parameters under different activity states to form a parameter set;

[0091] Identify the user's current activity status;

[0092] According to the user's current activity state, a heart rate range and heart rate variability parameter corresponding to the user's current activity state are selected from the parameter set as a criterion for determining heart rate abnormality;

[0093] Identify heart rate beats from purified physiological signals;

[0094] Compare heart rate pulses with the criteria for judging abnormal heart rate;

[0095] Based on the comparison results, it is determined whether there is an abnormal heart rate.

[0096] Among them, the heart rate variability parameter refers to a quantitative indicator of the change in the small time interval between heartbeat cycles. It can be measured by time domain indicators such as the standard deviation of adjacent heartbeat intervals, the root mean square of the difference between adjacent heartbeat intervals, or frequency domain indicators such as high-frequency power, low-frequency power, etc. Its purpose is to reflect the autonomic nervous system's ability to regulate cardiac activity; the activity state refers to the user's current physiological or physical situation, which can be identified by analyzing the user's motion sensor data (such as accelerometer, gyroscope data), geographic location information, or combining user input, etc., with the purpose of providing necessary contextual information for heart rate judgment; the parameter set refers to a database or model established for users in different activity states, containing corresponding heart rate ranges and heart rate variability parameters. It can be implemented using a pre-trained model, a user-personalized model updated through continuous learning, or a rule set set based on expert experience. Its purpose is to provide personalized and dynamically adjusted judgment basis for abnormal heart rate judgment.

[0097] The solution of this application constructs a parameter set tailored to individual user differences by acquiring and updating the user's heart rate range and heart rate variability parameters under different activity states. This parameter set can change over time and can continuously learn and adjust based on the user's actual physiological data, ensuring that the judgment criteria are timely and accurate. The system then identifies the user's current activity state, such as resting, walking, running, or sleeping. Based on the identified activity state, the system selects the heart rate range and heart rate variability parameters that match the current activity state from a pre-established parameter set, using these parameters as the basis for determining whether the current heart rate is abnormal. After obtaining the purified physiological signal, the system identifies the heart rate pulsation, which forms the basis for heart rate analysis. The identified heart rate pulsation is then compared with the judgment criteria selected based on the current activity state. This comparison not only considers whether the heart rate exceeds the normal range but also considers whether the heart rate variability conforms to the physiological characteristics of the current activity state. Finally, based on the comparison results, the system determines whether a heart rate abnormality exists. This series of steps forms a judgment process that can self-adjust and adapt to the user's context.

[0098] It is worth noting that the purified physiological signal used in this solution is obtained after the original physiological signal has been processed for artifact removal and continuous fluctuation signal suppression in the previous step. This preprocessing ensures the purity of the input signal and effectively reduces the adverse effects of motion artifacts, environmental noise, and other non-cardiac interference on heart rate judgment. Therefore, by making heart rate abnormality judgments based on pure physiological signals, this solution can accurately identify the true heart rate beats and make detailed personalized judgments on this basis, thereby effectively avoiding misjudgments and missed judgments due to signal contamination and improving the stability of heart rate abnormality judgments.

[0099] In some preferred embodiments, the specific process of judging heart rate abnormalities on purified physiological signals can be implemented as follows. First, in order to obtain and update the user's heart rate range and heart rate variability parameters in different activity states and form a parameter set, the system can continuously collect the user's purified physiological signals in different scenarios in daily life, and combine other sensor data to automatically identify the user's activity state, such as resting, light activity, moderate activity, strenuous exercise or sleep. For each activity state, the system can use historical data to establish a baseline model, for example, by calculating the mean and standard deviation of the heart rate in that state, as well as the distribution range of heart rate variability parameters (such as RMSSD). These baseline data can be stored in a local database and updated regularly based on new, confirmed normal physiological data, such as using a sliding average or exponentially weighted average method to smooth the update to ensure the dynamic adaptability of the parameter set.

[0100] Next, to identify the user's current activity state, the system analyzes data from the underwear's built-in motion sensors (such as a three-axis accelerometer and gyroscope) in real time. By extracting features from this sensor data, such as calculating activity intensity index and posture change rate, and inputting this data into a pre-trained activity classification model (such as a support vector machine or neural network), the system outputs a label for the user's current activity state.

[0101] Then, based on the user's current activity state, the system can query and select the heart rate range (for example, a resting heart rate range of 50-90 beats / minute) and the normal range of heart rate variability parameters (for example, a normal range of RMSSD at rest) that correspond to the activity state label from the previously constructed parameter set. These selected parameters serve as the current criteria for determining heart rate abnormality.

[0102] To identify heartbeats from the cleaned physiological signal, sophisticated signal processing algorithms can be used, such as peak detection-based methods. This method first performs a bandpass filter on the cleaned physiological signal to remove high-frequency noise and low-frequency baseline drift. Then, an adaptive threshold or morphological operation is applied to identify R-wave peaks in the signal, thereby determining the time of each heartbeat.

[0103] Next, to compare the heart rate beats with the heart rate anomaly criteria, the system can calculate the average heart rate and heart rate variability parameters within the current time window. For example, the system can calculate the average heart rate by counting the number of heart rate beats identified in the last 30 seconds and calculating the RMSSD value of these beat intervals. These calculated heart rate and RMSSD values ​​are then compared with the previously selected heart rate anomaly criteria. For example, it can determine whether the current average heart rate exceeds the selected range or whether the RMSSD value is below its normal lower limit.

[0104] Finally, based on the comparison results, a determination is made as to whether there is a heart rate anomaly. If any one or more of the current heart rate or heart rate variability parameters exceeds the corresponding judgment criteria, the system can determine that a heart rate anomaly exists and trigger a corresponding warning mechanism, such as vibration, flashing indicator lights, or sending a notification to a connected smart device.

[0105] The present application further proposes to obtain and update the heart rate range and heart rate variability parameters of the user in different activity states, and the steps of forming a parameter set include: obtaining the user's original mechanical impact signal, preprocessing the original mechanical impact signal, and obtaining a preprocessed mechanical impact signal; identifying the heart rate pulsation in the purified physiological signal; taking the time point of occurrence of the heart rate pulsation as the starting point, and within a preset time window after the starting point, identifying the characteristic event synchronized with the heart rate pulsation from the preprocessed mechanical impact signal; calculating the time difference between the time point of occurrence of the heart rate pulsation and the time point of occurrence of the characteristic event; judging whether the time difference falls within the time threshold range to confirm the authenticity of the heart rate pulsation; using the confirmed authenticity of the heart rate pulsation to obtain and update the user's heart rate range and heart rate variability parameters in different activity states to form a parameter set; wherein, the parameter set is limited to the corresponding data when the purified physiological signal is not determined to be an abnormal heart rate.

[0106] The raw mechanical shock signal refers to the raw data stream generated by the user's body movement or external physical contact and collected by a mechanical sensor. This can be implemented using a triaxial accelerometer, MEMS gyroscope, or piezoelectric force sensor. Preprocessing refers to performing a series of signal processing operations on the raw mechanical shock signal to remove noise, artifacts, or irrelevant components, improving signal quality and feature identifiability. This can be achieved using methods such as low-pass filtering, high-pass filtering, band-pass filtering, notch filtering, wavelet denoising, or moving average. The preset time window refers to a pre-set, finite time interval after the onset of the heartbeat, which is used to limit the search range for synchronous characteristic events in the mechanical shock signal. This can be implemented using a fixed time period, a time period dynamically adjusted based on the user's physiological characteristics, or a time period adjusted based on activity status. Characteristic events refer to identifiable signal patterns or peaks in the preprocessed mechanical shock signal that are synchronous or specifically correlated with the heartbeat in time. This can be implemented using signal peaks, valleys, specific waveform features, or signal energy mutation points. The time difference refers to the time interval between the occurrence of the heart rate pulse identified in the purified physiological signal and the occurrence of the synchronous characteristic event identified in the preprocessed mechanical impact signal. It can be implemented as the absolute time difference, relative time difference, or phase difference between the two events. The time threshold refers to an acceptable time range used to determine whether the time difference meets the preset synchronization relationship. Only when the time difference falls within this range is the heart rate pulse considered to be real. It can be implemented using a fixed numerical range, a range adjusted according to individual physiological differences, or a range dynamically adjusted according to activity intensity.

[0107] The solution of this application significantly improves the reliability of heart rate data by introducing a multimodal signal verification mechanism, thereby optimizing the accuracy of heart rate anomaly detection. Specifically, the solution first obtains the user's original mechanical shock signal and preprocesses it to obtain a high-quality preprocessed mechanical shock signal. This step lays the foundation for the subsequent accurate identification of characteristic events from the mechanical shock signal, ensuring signal purity. Simultaneously, the system identifies heart rate pulsations within the purified physiological signal and uses the time points of these pulsations as a time reference. Subsequently, starting with the time point of each heart rate pulsation, within a preset time window, the system searches and identifies characteristic events synchronized with the heart rate pulsation within the preprocessed mechanical shock signal. By limiting the time window, the search range can be effectively narrowed, improving the efficiency and accuracy of synchronous event identification. Based on this, the solution further calculates the time difference between the time point of the heart rate pulsation and the time point of the identified characteristic event. This time difference is a key indicator of the degree of synchronization between the two signals. The system then determines whether this calculated time difference falls within a preset time threshold. Only when the time difference meets this threshold condition is the heart rate pulse confirmed as a valid pulse. This multimodal signal-based time synchronization verification mechanism effectively eliminates false heart rate pulses caused by noise or artifacts, thereby ensuring the high authenticity of the heart rate pulse data used for subsequent analysis. Ultimately, only these verified heart rate pulses are used to obtain and update the user's heart rate range and heart rate variability parameters under different activity states, thereby forming the parameter set. Furthermore, to further ensure the purity and reliability of the parameter set, the construction and update of the parameter set is limited to data corresponding to cases where the physiological signal is not determined to be abnormal. In this way, this solution avoids the inclusion of interfered or abnormal heart rate data in the parameter set, thereby ensuring the accuracy and representativeness of the parameter set. Through this mechanism, this solution provides a more accurate and reliable judgment standard for subsequent heart rate anomaly determination. Compared with methods that rely solely on a single physiological signal to identify heart rate pulses and construct parameter sets, this solution effectively overcomes the problem of inaccurate heart rate pulse identification caused by the susceptibility of physiological signals to interference by incorporating mechanical shock signals for cross-validation. This multi-signal collaborative verification approach ensures that the acquired heart rate range and heart rate variability parameters are closer to the user's true physiological state, significantly improving the reliability of heart rate anomaly detection and reducing the risk of false positives and missed detections. This rigorous screening of heart rate pulse authenticity directly improves the quality of the heart rate anomaly detection criteria, substantially enhancing the performance of the entire heart rate monitoring system.

[0108] In some preferred embodiments, the present application is implemented as follows. First, a triaxial accelerometer integrated into the wearable device can be used to acquire the user's raw mechanical shock signal. This accelerometer can capture minute vibrations and shocks of the user's body in real time. Subsequently, the acquired raw mechanical shock signal is preprocessed. For example, a low-pass filter with a cutoff frequency of 20 Hz can be applied to remove high-frequency noise, and a moving average filter can be used to smooth the signal, thereby obtaining a preprocessed mechanical shock signal. Simultaneously, the system identifies heart rate pulses from the purified physiological signal, which has undergone artifact removal and suppression. This can be achieved by applying a sophisticated peak detection algorithm, such as the Pan-Tompkins algorithm, which accurately identifies the R-wave peak in the PPG signal and records its occurrence time. Next, using the R-wave peak time of each identified heart rate pulse as the starting point, the system searches for characteristic events synchronized with the heart rate pulse within a preset time window after the starting point, for example, within a range of 50 to 150 milliseconds. This characteristic event can specifically be a specific waveform peak in the mechanical shock signal associated with ventricular contraction. For example, after a heartbeat occurs, the impact of blood ejection on the blood vessel wall may produce a small, identifiable vibration peak in the mechanical shock signal. The system then calculates the time difference between the peak of the heartbeat's R wave and the peak of the identified mechanical shock signal characteristic event. For example, if the R wave peak occurs at time T1 and the peak of the mechanical shock characteristic event occurs at time T2, the time difference is T2-T1. Furthermore, the system determines whether this calculated time difference falls within a preset time threshold, for example, a time threshold set between -20 milliseconds and +20 milliseconds. If the time difference falls within this range, the heartbeat is confirmed to be authentic; otherwise, the heartbeat is likely an artifact or noise and is discarded. Finally, only these heartbeats that have been verified to be authentic are used to obtain and update the user's heart rate range and heart rate variability parameters during different activity states, forming a parameter set. For example, when the user is at rest, the system collects a large amount of verified resting heart rate and pulse data and calculates its average heart rate and heart rate variability (such as SDNN or RMSSD) as parameters in the resting state. When the user enters the exercise state, the system switches to the parameter update mode in the exercise state. It should be noted that only when the purified physiological signal is not judged as an abnormal heart rate will the corresponding heart rate and pulse data be included in the update of the parameter set, thereby avoiding the contamination of the normal parameter set by abnormal data.

[0109] The present application further proposes that the steps of performing suppression processing on the continuous fluctuation signal component include:

[0110] Perform frequency analysis on the purified physiological signals to extract the dominant frequency characteristics and energy distribution range of the continuous fluctuation signal components;

[0111] Configure the filter parameters based on the dominant frequency characteristics and energy distribution range of the continuously fluctuating signal components;

[0112] The configured filter is applied to purify the physiological signal to suppress the continuously fluctuating signal components.

[0113] Frequency analysis refers to the process of converting a time-domain signal into a frequency-domain signal through mathematical transformations. This can be achieved using methods such as the fast Fourier transform, short-time Fourier transform, or wavelet transform. Its purpose is to reveal the strength and distribution of different frequency components in the signal. The dominant frequency characteristics of a continuously fluctuating signal component refer to the frequency point or frequency range where the energy of the continuously fluctuating signal component is most concentrated in the frequency analysis results. This can be expressed as the peak frequency in a spectrum or the average frequency within a specific frequency band. Its purpose is to determine the center frequency of the noise that needs to be suppressed. The energy distribution range refers to the range occupied by the continuously fluctuating signal component in the frequency domain and the distribution of its energy across different frequencies. This can be expressed as the bandwidth of the spectrum, an energy decay curve, or the energy percentage of a specific frequency band. Its purpose is to fully understand the degree of noise diffusion in the frequency domain and provide a basis for the precise configuration of subsequent filter parameters. Configuring filter parameters refers to dynamically adjusting filter settings based on the frequency characteristics of the signal. Specifically, this can include adjusting the filter's cutoff frequency, center frequency, bandwidth, order, or filter type. Its purpose is to enable the filter to specifically suppress the target noise while maximally retaining the effective signal.

[0114] The solution of this application accurately identifies the frequency characteristics of the continuously fluctuating signal component by performing frequency analysis on the purified physiological signal. Specifically, by performing frequency analysis on the purified physiological signal, the dominant frequency characteristics and energy distribution range of the continuously fluctuating signal component can be extracted. These frequency characteristics and energy distribution information serve as the frequency domain "fingerprint" of the continuously fluctuating signal component, revealing the frequency center and diffusion range of this non-pulsatile noise. Based on these frequency characteristics and energy distribution ranges acquired in real time, the system intelligently configures the filter parameters. This configuration does not rely on fixed preset values, but rather adaptively adjusts based on the actual frequency characteristics of the continuously fluctuating signal component in the current signal. For example, the filter's cutoff frequency, bandwidth, or stopband attenuation can be adjusted to ensure that the filter accurately covers and attenuates the target noise frequency. This finely configured filter is then applied to the purified physiological signal to effectively suppress the continuously fluctuating signal component. The synergistic effect of this series of steps is that it overcomes the limitations of traditional fixed filters in processing continuously fluctuating signal components that partially overlap with the heart rate frequency range. Based on removing physiological artifact signals from the original physiological signal and obtaining a purified physiological signal, this solution further provides a dynamic and adaptive suppression mechanism for the continuously fluctuating signal components in the purified physiological signal that are difficult to remove with simple artifact removal. By performing real-time frequency analysis on the noise and adjusting the filter accordingly, it is possible to avoid distortion of the effective heart rate pulse signal caused by excessive filtering, while effectively removing noise that is aliased with the heart rate signal. This method enables a purer heart rate signal to be obtained even in complex noise environments, thereby providing a high-quality data foundation for subsequent heart rate abnormality judgments and significantly improving the accuracy and reliability of heart rate detection.

[0115] In some preferred embodiments, a fast Fourier transform (FFT) algorithm can be used to perform frequency analysis on the purified physiological signal. For example, a 256-point data window can be taken every second and subjected to FFT processing to obtain the spectral information of the signal within that time window. From this spectral information, the frequency peak of the continuously fluctuating signal component can be identified. For example, if a significant energy peak is detected between 5Hz and 10Hz, this frequency can be determined as the dominant frequency characteristic of the continuously fluctuating signal component. The energy distribution around this peak can also be analyzed. For example, the frequency range where the energy on either side of the peak drops to half of the peak can be calculated to determine its energy distribution range. Based on these extracted dominant frequency characteristics and energy distribution range, an adaptive band-stop filter can be configured. For example, if the dominant frequency characteristic is 8Hz and the energy distribution range is 7Hz to 9Hz, the center frequency of the band-stop filter can be set to 8Hz and the bandwidth to 2Hz. The filter type can be a Butterworth filter or a Chebyshev filter, and its order can be dynamically adjusted based on the suppression requirements. For example, a 3rd-order or 4th-order filter can be selected to achieve the desired attenuation effect based on the noise intensity and bandwidth. Finally, the configured adaptive band-stop filter is applied to the purified physiological signal. This means that each data point of the purified physiological signal is processed by this filter, which is adjusted according to the real-time noise characteristics. Through this processing, the energy of the continuously fluctuating signal components in the 7Hz to 9Hz range is significantly attenuated, while the heart rate pulse signal is preserved, resulting in a purer heart rate signal with the continuously fluctuating signal components effectively suppressed.

[0116] The present application further proposes monitoring the integrity of the pulsation waveform in the purified physiological signal during the process of suppressing the continuous fluctuation signal component, and reducing the suppression strength of the filter when the pulsation waveform distortion is detected.

[0117] Among them, monitoring the integrity of the pulsating waveform in the purified physiological signal refers to real-time or periodic evaluation of whether the pulsating waveform in the purified physiological signal maintains its expected shape and characteristics. It can be achieved by waveform feature extraction and threshold comparison, template matching or machine learning model, etc., with the purpose of timely discovering the negative impact of filtering on the effective physiological signal; pulsating waveform distortion refers to the change in the key features of the pulsating waveform in the purified physiological signal after filtering, such as peak amplitude, waveform width and rising edge slope, which significantly deviate from the normal or expected state. It can be manifested as the waveform being flattened, narrowed, reduced in amplitude or abnormal spikes, etc., with the purpose of identifying the damage to the signal quality caused by excessive filtering; reducing the suppression strength of the filter refers to reducing the attenuation degree of the filter on the target frequency component, so that the attenuation effect of the filter on the signal is weakened. It can be achieved by reducing the filter coefficient of the filter, adjusting the cutoff frequency or bandwidth of the filter, or switching to a filter type with lower suppression strength, etc., with the purpose of reducing the negative impact on the pulsating waveform and restoring its integrity.

[0118] The solution of this application achieves adaptive control of the filtering process by introducing a real-time monitoring mechanism for the integrity of the pulsating waveform in the purified physiological signal during the process of suppressing the continuously fluctuating signal component. Specifically, while the filter is configured based on the dominant frequency characteristics and energy distribution range of the continuously fluctuating signal component and applied to the purified physiological signal to suppress interference, the system synchronously and continuously evaluates the pulsating waveform in the filtered purified physiological signal. Once distortion of the pulsating waveform is detected, such as abnormal attenuation of the waveform amplitude, distortion of the waveform shape, or key characteristic parameters exceeding the preset range, this indicates that the current filter suppression strength may be too high, causing unnecessary damage to the pulsating waveform. At this time, the system immediately responds and dynamically reduces the filter suppression strength, for example by adjusting the filter gain, bandwidth, or attenuation coefficient to reduce its attenuation effect on the signal. This dynamic adjustment mechanism enables the filter to effectively suppress the continuously fluctuating signal component while avoiding excessive damage to the heart rate pulsating waveform, thereby achieving a balance between noise suppression and effective signal preservation. Thanks to this real-time feedback and adaptive adjustment, this solution effectively addresses the potential for pulse waveform distortion that can occur with traditional fixed-parameter filtering or filtering methods based solely on frequency analysis, thereby ensuring accurate and reliable heart rate detection. This approach, combined with previous solutions that configure filter parameters based on frequency analysis, creates a more robust and intelligent signal processing pipeline that better adapts to complex and changing physiological signal environments and avoids the loss or misjudgment of true heart rate information due to excessive filtering.

[0119] In some preferred embodiments, when filtering purified physiological signals to suppress persistent fluctuating signal components, the integrity of the pulsation waveform in the purified physiological signals can be monitored and adjusted in the following manner. First, each heart rate pulsation waveform can be identified in real time from the purified physiological signals, and key pulsation characteristic parameters, such as peak amplitude, waveform width (e.g., half-width), and rising edge slope, can be extracted. These parameters can reflect the energy, duration, and steepness of the pulsation waveform. Next, a set of reference ranges for the pulsation characteristic parameters can be preset. This reference range can be obtained through statistical analysis of a large amount of normal physiological signal data, or personalized based on the user's historical health data. During the filtering process, the currently extracted pulsation characteristic parameters are compared with the preset reference ranges. If any one or more pulsation characteristic parameters exceed their corresponding reference ranges, such as a significant decrease in peak amplitude, an abnormally narrowed waveform width, or a significant decrease in rising edge slope, the pulsation waveform can be determined to be distorted. Once pulsation waveform distortion is detected, the system can immediately trigger a feedback mechanism to reduce the suppression strength of the current filter. Specifically, if a digital filter is used, the filter's gain coefficient can be reduced or its cutoff frequency adjusted to minimize attenuation of continuously fluctuating signal components. For example, for a band-stop filter, its passband can be appropriately widened or its stopband attenuation depth reduced. This dynamic adjustment ensures that the original form of the heart rate waveform is preserved to the greatest extent possible while effectively removing interference, avoiding loss or misinterpretation of heart rate information due to excessive filtering.

[0120] The present application further proposes that the steps of monitoring the integrity of the pulsating waveform in the purified physiological signal include:

[0121] Identify heart rate beats from purified physiological signals;

[0122] Extracting pulse characteristic parameters from the heart rate pulse, the pulse characteristic parameters including peak amplitude, waveform width and rising edge slope;

[0123] Get the user's current wearing status information;

[0124] Dynamically adjust the reference range of pulsation characteristic parameters based on the user's current wearing status information;

[0125] comparing the beat characteristic parameter with a reference range of the adjusted beat characteristic parameter;

[0126] Based on the comparison result, the integrity of the beat waveform in the purified physiological signal is determined.

[0127] Among them, identifying heart rate pulsation refers to locating and distinguishing each independent heart rate pulsation cycle from a continuous purified physiological signal stream. It can be achieved by using technologies such as peak detection, template matching or machine learning classification, and its purpose is to provide analysis objects for subsequent waveform feature extraction; pulsation feature parameters refer to key values ​​that can quantitatively describe the shape of the heart rate pulsation waveform, specifically the peak amplitude, waveform width and rising edge slope, which can be calculated by using signal processing algorithms for the identified heart rate pulsation waveform, and its purpose is to provide multi-dimensional data to evaluate the integrity of the waveform; the user's current wearing status information refers to data reflecting the contact status and relative position between the wearable device and the user's body, which can be collected by built-in sensors (such as accelerometers, gyroscopes, pressure sensors or strain sensors), information manually entered by the user, or through The purpose of obtaining the pulse characteristic parameters by image recognition and other means is to provide a basis for the subsequent dynamic adjustment of the reference range of the pulse characteristic parameters; the dynamic adjustment of the reference range of the pulse characteristic parameters refers to the real-time correction of the normal threshold interval of each characteristic parameter used to judge the integrity of the pulse waveform according to the changes in the user's wearing status. It can be implemented by a preset lookup table, a statistical model based on historical data or an adaptive algorithm. Its purpose is to improve the accuracy and adaptability of the waveform integrity judgment; determining the integrity of the pulse waveform in the purified physiological signal refers to a comprehensive assessment of whether the current heart rate pulse waveform maintains its inherent, undistorted form based on the comparison results of the pulse characteristic parameters and the adjusted reference range. It can be implemented by rule-based logical judgment, fuzzy reasoning system or machine learning classifier. Its purpose is to provide a decision basis for whether the filter suppression strength needs to be adjusted.

[0128] The solution of this application achieves refined monitoring of the integrity of the pulsation waveform in the purified physiological signal through a series of logically rigorous steps, thereby effectively avoiding excessive impact on the true heart rate pulsation waveform while suppressing the continuous fluctuation signal components. First, each independent heart rate pulsation is identified from the purified physiological signal. This is the basis for all subsequent waveform analysis and ensures the focus on the valid signal. On this basis, multi-dimensional pulsation feature parameters are extracted from these identified heart rate pulsations, including peak amplitude, waveform width and rising edge slope. These parameters can comprehensively quantify the morphological characteristics of the pulsation waveform and provide a quantitative basis for determining whether it is distorted. Considering the inherent influence of the wearing state of the wearable device on the waveform morphology of the physiological signal, this solution further obtains the user's current wearing state information and dynamically adjusts the reference range of the pulsation feature parameters based on this information. This dynamic adjustment mechanism enables the judgment of waveform integrity to adapt to different wearing conditions, avoids misjudgment due to changes in wearing state, and significantly improves the accuracy and adaptability of the judgment. Subsequently, the extracted pulsation characteristic parameters are compared with the dynamically adjusted reference range. Through this comparison, it is possible to accurately evaluate whether the current pulsation waveform has deviated from its normal form. Finally, based on the comparison results, the system can determine the integrity of the pulsation waveform in the purified physiological signal. When it is determined that the waveform is distorted, combined with the aforementioned scheme, the system can promptly reduce the suppression strength of the filter, thereby avoiding irreversible damage to the true heart rate pulsation waveform caused by excessive filtering. It is precisely because of this refined and adaptive waveform integrity monitoring mechanism that while effectively suppressing the continuous fluctuation signal components, the original form of the heart rate pulsation waveform can be retained to the maximum extent, ensuring the accuracy and reliability of heart rate detection.

[0129] In some preferred embodiments, the present application is implemented as follows: When monitoring the integrity of the pulsation waveform in a purified physiological signal, a peak detection algorithm based on an adaptive threshold can be first used to identify heart rate pulses. For example, the peak detection threshold can be dynamically set by calculating the local mean and standard deviation of the signal to adapt to changes in signal amplitude. Once the heart rate pulses are identified, pulse characteristic parameters can be extracted from each pulse waveform. Specifically, the peak amplitude can be defined as the vertical distance between the peak of the pulse waveform and the baseline; the waveform width can be defined as the duration at half the peak of the waveform (full width at half maximum); and the rising edge slope can be calculated as the average slope of the waveform from the baseline to 80% of the peak. To obtain information about the user's current wearing state, a triaxial accelerometer and strain sensor integrated into the wearable bra can be utilized. For example, accelerometer data can be used to identify the user's arm movement pattern and posture, while the strain sensor can measure the tightness of the bra's contact with the skin. Based on this wearing state information, the reference range of the pulse characteristic parameters can be dynamically adjusted. For example, a lookup table can be preset that stores normal reference ranges for peak amplitude, waveform width, and rising edge slope based on different wearing states (e.g., tight, moderate, loose) and exercise intensities (e.g., resting, light activity, vigorous exercise). When the system identifies the current wearing state as "relaxed" and the user as "resting," the reference lower limit for peak amplitude can be appropriately relaxed to avoid misclassifying normal but low-amplitude pulses as distortion. Subsequently, the currently extracted pulse feature parameters are compared with the adjusted reference ranges obtained from the lookup table. For example, if the peak amplitude is below the adjusted reference lower limit, or the waveform width exceeds the adjusted reference range, potential waveform distortion is considered. Finally, a weighted scoring-based decision mechanism can be employed to determine the integrity of the pulse waveform in the purified physiological signal. For example, a weight is assigned to each pulse feature parameter that exceeds the reference range, and a total score is calculated. If the total score exceeds a preset distortion threshold, the pulse waveform is determined to be distorted, triggering an operation to reduce the filter suppression strength.

[0130] This application further proposes that the steps of dynamically adjusting the reference range of the pulsation characteristic parameters according to the user's current wearing status information include:

[0131] Collect inertial measurement data and strain data at multiple locations in the underwear;

[0132] Determine the user's current wearing status based on a preset wearing status recognition model;

[0133] The corresponding wearing status label is output according to the user's current wearing status, and the wearing status label is used for adjusting the reference range of the pulsation characteristic parameters.

[0134] Among them, inertial measurement data refers to data reflecting the motion state of an object in space, which may specifically include information such as acceleration, angular velocity and posture, and its purpose is to capture the dynamic behavior of the worn underwear, such as shaking, displacement or relative motion; strain data refers to data reflecting the degree of deformation of the object under force, which may specifically include deformation information such as stretching, compression or bending, and its purpose is to quantify the fit between the worn underwear and the body, such as whether it is too loose or too tight; the preset wearing state recognition model refers to a trained and optimized algorithm or system, which may specifically adopt a machine learning model such as a support vector machine, a neural network or a decision tree, and its purpose is to automatically identify and classify different wearing states, such as good wearing, loose wearing or too tight wearing, based on the input inertial measurement data and strain data; the wearing state label refers to the classification result that provides a general description of the identified wearing state, which may specifically be a string, enumeration value or numerical code, and its purpose is to simplify and standardize complex wearing state information so that subsequent modules can directly use the label to adjust the reference range of the pulsation characteristic parameters.

[0135] The solution of this application collects inertial measurement data and strain data from multiple locations within the bra to comprehensively capture multi-dimensional information related to the wear state. Inertial measurement data can reflect the bra's motion state, such as whether it's loose or shifting, while strain data can reflect the fit of the bra against the body, such as whether it's too tight or too loose. By collecting data from multiple locations, a more complete description of the bra's wear state can be achieved, avoiding the limitations of a single data source. Based on this multi-source data, the system utilizes a preset wear state recognition model to determine the user's current wear state. This model is pre-trained and can accurately identify different wear states, such as good fit, loose fit, or too tight, based on the input inertial measurement data and strain data. This model-based judgment method avoids errors caused by subjective judgment and improves the accuracy of wear state recognition. The system then outputs a corresponding wear state label based on the user's current wear state. These labels provide a summary description of the wear state recognition results, such as "good fit," "loose fit," or "too tight." These labels serve as key information and are provided to the reference range adjustment module for the pulsation characteristic parameters. Because this solution accurately captures the user's current wearing state information and converts it into a wear state tag that the system can utilize, it enables dynamic and precise adjustment of the reference range of pulsation characteristic parameters. For example, when the wearing state is "loose," the system can appropriately widen the reference range of the pulsation characteristic parameters based on this tag to avoid misjudging normal pulsations as abnormal. Conversely, when the wearing state is "too tight," corresponding adjustments can also be made. This precise wearing state identification and reference range adjustment mechanism is closely integrated with the process of monitoring the integrity of the pulsation waveform in the purified physiological signal. When monitoring the integrity of the pulsation waveform, the extracted pulsation characteristic parameters must be compared with a reference range. If this reference range does not accurately reflect the user's current wearing state, it may lead to a misjudgment of the pulsation waveform integrity, which in turn affects the subsequent adjustment of the filter suppression strength and the ultimate determination of heart rate abnormalities. Through this solution, the reference range of the pulsation characteristic parameters can be dynamically optimized based on the actual wearing state, making the determination of pulsation waveform integrity more reliable. For example, if bras are worn too loosely, the pulsation signal may experience reduced amplitude or waveform distortion. Failure to adjust the reference range could lead to misinterpretation as pulsation waveform distortion, unnecessarily reducing the filter's suppression strength and failing to effectively suppress continuously fluctuating signal components. Conversely, if bras are worn too tightly, the pulsation signal may be suppressed, also resulting in misinterpretation if the reference range is not adjusted.Therefore, this solution provides accurate wearing status information, making the reference range adjustment of the pulsation characteristic parameters more reasonable, thereby improving the accuracy of the integrity judgment of the pulsation waveform, and then optimizing the suppression effect of the continuous fluctuation signal component, and ultimately improving the accuracy and reliability of the entire heart rate detection method for abnormal heart rate judgment, effectively solving the problem of inaccurate heart rate abnormality judgment caused by uncertainty in wearing status.

[0136] In some preferred embodiments, a wearable bra can integrate multiple miniature inertial measurement units (IMUs) and flexible strain sensors. For example, IMUs can be deployed at key locations on the bra, such as the chest, flanks, and shoulder straps. These IMUs can utilize a combination of triaxial accelerometers and triaxial gyroscopes to collect linear acceleration and angular velocity data in different directions. Furthermore, flexible strain sensors can be embedded in key areas of the bra where it contacts the skin, such as below the sternum, along the sides of the ribs, and on the shoulders. These sensors can utilize piezoresistive or capacitive principles to monitor the deformation of the bra material under load in real time, thereby reflecting the fit of the bra. The collected inertial measurement and strain data can be transmitted to an embedded processor, such as a low-power microcontroller or a dedicated signal processing chip. This processor can pre-store a wearing state recognition model. This model can be a deep learning-based convolutional neural network or recurrent neural network, which is trained offline using a large amount of inertial measurement and strain data from different wearing states. For example, data of users in various states such as good wearing, loose wearing, too tight wearing, underwear shifting, etc. can be collected and labeled to train the model to recognize these states. When real-time data is input, the model extracts and classifies the data features and outputs corresponding wearing state labels, such as "good wearing", "slightly loose", "severely loose", "too tight wearing" or "partially shifted". These wearing state labels can then be passed as input parameters to the reference range adjustment module of the pulsation characteristic parameters. The module can preset a lookup table or a parameter adjustment algorithm to dynamically adjust the upper and lower reference values ​​of the pulsation characteristic parameters (such as peak amplitude, waveform width and rising edge slope) according to the received wearing state labels. For example, when the label is "severely loose", the amplitude of the pulsation waveform may be generally low. At this time, the reference lower limit of the peak amplitude can be appropriately lowered to avoid misjudging normal low-amplitude pulsations as distortion.

[0137] refer to Figure 2 This application further proposes a heart rate detection system for wearing underwear, which is applied to a heart rate detection method for wearing underwear. The system includes:

[0138] an information processing module, configured to obtain the user's motion information and identify specific body posture changes based on the user's motion information;

[0139] A physiological artifact signal generation module, which generates a physiological artifact signal related to a specific body posture change based on the specific body posture change;

[0140] A physiological signal purification module is used to remove physiological artifact signals from original physiological signals to obtain purified physiological signals;

[0141] The heart rate abnormality judgment module judges the heart rate abnormality of the purified physiological signal.

[0142] Among them, the information processing module refers to the unit responsible for receiving and analyzing user motion data. It can be implemented by a microcontroller or a dedicated digital signal processor integrated with inertial sensors such as accelerometers and gyroscopes. Its purpose is to sense the user's activity status and identify specific body posture changes, providing context information for subsequent signal processing; the physiological artifact signal generation module refers to the unit that constructs or predicts the corresponding interference signal based on the identified body posture changes. It can be implemented by a software module based on a pre-trained model or algorithm. The model or algorithm can generate a simulated physiological artifact waveform based on the posture change parameters. Its purpose is to provide an accurate artifact reference for subsequent signal purification; the physiological signal purification module refers to the unit that separates and removes interference components from the original physiological signal. It can be implemented by hardware or software using a digital filter, adaptive noise elimination algorithm or blind source separation algorithm. Its purpose is to improve the purity of the heart rate signal and ensure the accuracy of subsequent analysis; the heart rate abnormality judgment module refers to the unit that analyzes the processed physiological signal to identify potential heart rate abnormalities. It can be implemented by a software module based on a rule engine, machine learning classifier or statistical analysis model. Its purpose is to provide real-time health warnings for users.

[0143] The solution of the present application concretizes the key steps in the heart rate detection method into independent system modules, thereby constructing a system that can stably and reliably perform heart rate detection tasks. Specifically, the information processing module first obtains the user's motion information and identifies specific body posture changes based on this information. These posture change information is then passed to the physiological artifact signal generation module, which uses this information to generate physiological artifact signals related to specific body posture changes. These generated artifact signals are not directly used for heart rate judgment, but are used as a reference or model to guide the work of the physiological signal purification module. The physiological signal purification module receives the original physiological signal and, in combination with the artifact information provided by the physiological artifact signal generation module, accurately removes these physiological artifact signals from the original physiological signal, thereby obtaining a purer physiological signal. Finally, the heart rate abnormality judgment module analyzes these purified physiological signals to identify whether there is a heart rate abnormality. It is precisely because of this modular design that each functional unit has clear responsibilities and can be optimized for its respective tasks. For example, the information processing module can focus on the acquisition of motion data and the accuracy of posture recognition, the physiological artifact signal generation module can focus on the precise construction of the artifact model, the physiological signal purification module can focus on efficient artifact removal, and the heart rate abnormality judgment module can focus on the accurate analysis of the purified signal. This system architecture with division of labor and cooperation enables the entire heart rate detection process to effectively cope with the interference caused by human activities and changes in wearing status in dynamic and complex application scenarios such as wearing underwear, ensuring the stability and accuracy of heart rate detection. Compared with simply describing the method steps, this system ensures the effective implementation of the method steps and the close coordination between the various functions by providing a specific execution carrier, thereby overcoming the problem that a simple method cannot guarantee the execution effect and significantly improving the reliability of heart rate monitoring.

[0144] In some preferred embodiments, the present application is implemented as follows: The wearable underwear heart rate detection system can be integrated onto a microcircuit board and embedded in a specific location within the wearable underwear. The information processing module can be a low-power microcontroller, such as an ARM Cortex-M series microcontroller, integrating a multi-axis accelerometer and gyroscope. The microcontroller continuously collects the user's motion data and analyzes this data in real time using a built-in algorithm to identify current body posture, such as switching from sitting to standing, arm raising, or torso twisting. The physiological artifact signal generation module can be a software program running on the microcontroller that includes a lightweight artifact prediction model. When the information processing module identifies a specific body posture change, the model generates a simulated physiological artifact signal based on the type and intensity of the posture change. For example, if a rapid arm raising is detected, the module can generate a simulated artifact waveform with a specific frequency and amplitude that simulates the PPG signal fluctuations caused by changes in venous return. The physiological signal purification module can be a digital signal processing algorithm, also running on the microcontroller. The algorithm receives the raw physiological signal from the PPG sensor and combines it with the simulated artifact signal provided by the physiological artifact signal generation module. It uses adaptive filtering technology, such as the least mean square (LMS) algorithm, to remove artifact components from the raw signal in real time. In this way, non-cardiac interference such as motion artifacts and respiratory artifacts can be effectively suppressed to obtain a clear heart rate signal. The heart rate abnormality judgment module can be specifically a judgment logic based on rules or machine learning, running on the same microcontroller. This module performs waveform analysis on the purified physiological signal, extracts parameters such as heart rate beat period and heart rate variability, and compares them with the preset or dynamically updated user health baseline. Once the heart rate is detected to be out of the normal range or an abnormal pattern occurs, the system can trigger an alarm or record the abnormal event, and send the information to the user's smartphone or other display device through a wireless communication module (such as a Bluetooth low power module).

[0145] The contents disclosed above are only preferred feasible embodiments of the present invention and do not limit the scope of protection of the present invention. Therefore, all equivalent technical changes made using the contents of the present invention description and drawings are included in the scope of protection of the present invention. In addition, the elements therein can be updated as technology develops.

Claims

1. A method for detecting heart rate while wearing underwear, characterized in that: The method comprises the following steps: Acquiring the user's motion information and identifying specific body posture changes based on the user's motion information; generating a physiological artifact signal related to the specific body posture change based on the specific body posture change; Removing physiological artifact signals from original physiological signals to obtain purified physiological signals; Determine abnormal heart rate based on purified physiological signals.

2. A method for detecting heart rate while wearing underwear as claimed in claim 1, characterized in that: The steps of generating a physiological artifact signal related to the specific body posture change based on the specific body posture change include: extracting dynamic parameters related to specific body posture changes from the user's motion information; Based on specific body posture changes and corresponding dynamic parameters, a physiological artifact signal is generated through an artifact prediction model, wherein at least one of the morphology, amplitude and duration of the physiological artifact signal is dynamically adjusted according to the dynamic parameters.

3. The method for detecting heart rate while wearing underwear according to claim 1, wherein: After removing the physiological artifact signal from the original physiological signal to obtain the purified physiological signal, the method further includes: Identify and purify the continuous fluctuation signal component in the physiological signal, where the continuous fluctuation signal component is a non-pulsatile signal, and the frequency of the continuous fluctuation signal component partially overlaps with the heart rate frequency range; Suppression processing is performed on continuously fluctuating signal components.

4. The method for detecting heart rate while wearing underwear according to claim 1, wherein: The steps for determining abnormal heart rate based on purified physiological signals include: Obtain and update the user's heart rate range and heart rate variability parameters under different activity states to form a parameter set; Identify the user's current activity status; According to the user's current activity state, a heart rate range and heart rate variability parameter corresponding to the user's current activity state are selected from the parameter set as a criterion for determining heart rate abnormality; Identify heart rate beats from purified physiological signals; Compare heart rate pulses with the criteria for judging abnormal heart rate; Based on the comparison results, it is determined whether there is an abnormal heart rate.

5. A method for detecting heart rate while wearing underwear as claimed in claim 4, characterized in that: The steps of obtaining and updating the heart rate range and heart rate variability parameters of the user in different activity states and forming a parameter set include: Acquire the user's original mechanical shock signal, pre-process the original mechanical shock signal, and obtain a pre-processed mechanical shock signal; Identify heart rate beats in purified physiological signals; Taking the time point of occurrence of the heart beat as the starting point, within a preset time window after the starting point, characteristic events synchronized with the heart beat are identified from the pre-processed mechanical shock signal; Calculating the time difference between the occurrence time of the heart beat and the occurrence time of the characteristic event; Determine whether the time difference falls within the time threshold to confirm the authenticity of the heart rate beat; The confirmed authenticity of the heart rate pulse is used to obtain and update the user's heart rate range and heart rate variability parameters under different activity states to form a parameter set; The parameter set is limited to data corresponding to the case where the purified physiological signal is not determined to be an abnormal heart rate.

6. The method for detecting heart rate while wearing underwear according to claim 3, wherein: The steps of suppressing the continuously fluctuating signal component include: Perform frequency analysis on the purified physiological signals to extract the dominant frequency characteristics and energy distribution range of the continuous fluctuation signal components; Configure the filter parameters based on the dominant frequency characteristics and energy distribution range of the continuously fluctuating signal components; The configured filter is applied to purify the physiological signal to suppress the continuously fluctuating signal components.

7. A method for detecting heart rate while wearing underwear as claimed in claim 6, characterized in that: In the process of suppressing the continuous fluctuation signal component, the integrity of the beat waveform in the purified physiological signal is monitored, and when the beat waveform distortion is detected, the suppression strength of the filter is reduced.

8. A method for detecting heart rate while wearing underwear as claimed in claim 7, characterized in that: The steps for monitoring the integrity of the beat waveform in the cleaned physiological signal include: Identify heart rate beats from purified physiological signals; Extracting pulse characteristic parameters from the heart rate pulse, the pulse characteristic parameters including peak amplitude, waveform width and rising edge slope; Get the user's current wearing status information; Dynamically adjust the reference range of pulsation characteristic parameters based on the user's current wearing status information; comparing the beat characteristic parameter with a reference range of the adjusted beat characteristic parameter; Based on the comparison result, the integrity of the beat waveform in the purified physiological signal is determined.

9. A method for detecting heart rate while wearing underwear as claimed in claim 8, characterized in that: The steps of dynamically adjusting the reference range of the pulsation characteristic parameter according to the user's current wearing state information include: Collect inertial measurement data and strain data at multiple locations in the underwear; Determine the user's current wearing status based on a preset wearing status recognition model; The corresponding wearing status label is output according to the user's current wearing status, and the wearing status label is used for adjusting the reference range of the pulsation characteristic parameters.

10. A heart rate detection system for wearing underwear, applied to the heart rate detection method for wearing underwear according to claim 1, characterized in that: The system includes: an information processing module, configured to obtain the user's motion information and identify specific body posture changes based on the user's motion information; A physiological artifact signal generation module, which generates a physiological artifact signal related to a specific body posture change based on the specific body posture change; A physiological signal purification module is used to remove physiological artifact signals from original physiological signals to obtain purified physiological signals; The heart rate abnormality judgment module judges the heart rate abnormality of the purified physiological signal.

Images