Self-adaptive multi-scale envelope peak point BCG heart rate detection method and system

Abstract

The invention provides a self-adaptive multi-scale envelope peak point BCG heart rate detection method and system, and the method comprises the steps: S1, collecting BCG signal data, and carrying out the wavelet processing of the collected BCG signal data through a self-adaptive threshold value selection method; s2, envelope line extraction is carried out on the BCG signal data after wavelet processing; s3, carrying out low-pass filtering processing on the extracted envelope line; s4, performing peak point searching on the signal after low-pass filtering by adopting an automatic multi-scale peak value searching method; and S5, solving an average peak point interval of the searched peak points, and calculating a heart rate value according to the interval. By combining the self-adaptive threshold and the wavelet transform technology, the heart rate signal can be extracted more effectively, and the problem that J peak recognition is unstable possibly occurring on complex and changeable BCG signals through a traditional fixed threshold method is avoided; by using envelope line extraction and adaptive peak detection technologies, the accuracy of heart rate monitoring is further enhanced, and the interference of noise and motion artifacts is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of biological signal processing, and in particular to an adaptive multi-scale envelope peak BCG heart rate detection method and system. Background Art

[0002] In the field of heart rate monitoring, especially in contactless monitoring, accurately extracting and calculating heart rate is a challenging task. Traditional contactless monitoring equipment usually uses a fixed threshold method to calculate heart rate, but this method often has difficulty in obtaining a stable J peak recognition effect on complex and changeable BCG (Ballistocardiogram) signals. BCG signals are a signal generated by measuring the tiny body movements caused by the beating of the heart. It can reflect the activity of the heart and is therefore widely used in heart rate monitoring.

[0003] The fixed threshold method has many limitations when processing BCG signals. First, due to the complexity and variability of BCG signals, fixed thresholds are often difficult to adapt to the signal characteristics of different subjects and in different environments, resulting in a decrease in the accuracy of heart rate calculation. Second, the fixed threshold method is highly dependent on the heartbeat template during the training phase. Once the wrong template is learned, it will affect all subsequent heartbeat detections. In addition, when the position or state of the subject changes, this method usually needs to be retrained to adapt to the new signal characteristics, which brings inconvenience in practical applications.

[0004] In order to solve the above problems, some improved methods have appeared in the prior art. For example, a patent of the People's Republic of China with application number CN202210651400.3 discloses a method for feature extraction and heart rate calculation of BCG signals in a complex environment. This method obtains multiple independent signal components through signal preprocessing and signal separation technology, and then combines signal feature matching with multiple feature fusion methods to identify the I, J, K, and L peaks of the BCG signal, and finally calculates the real-time heart rate based on the position of continuous peaks. Although this method can improve the accuracy of heart rate calculation to a certain extent, it requires a complex parameter system, has high requirements on the hardware resources of embedded instruments and equipment, and still relies on the establishment of a heartbeat template.

[0005] Therefore, in practical applications, how to further improve the accuracy of heart rate monitoring while reducing the requirements for hardware resources and reducing the dependence on heartbeat templates has become a problem that needs to be solved urgently. Summary of the invention

[0006] In view of the defects in the prior art, the object of the present invention is to provide an adaptive multi-scale envelope peak BCG heart rate detection method and system.

[0007] According to an adaptive multi-scale envelope peak BCG heart rate detection method provided by the present invention, the method comprises the following steps: Step S1: Collect BCG signal data, and perform wavelet processing on the collected BCG signal data in combination with an adaptive threshold selection method; Step S2: extracting the envelope of the BCG signal data after wavelet processing; Step S3: performing low-pass filtering on the extracted envelope; Step S4: using an automatic multi-scale peak search method to search for peak points of the low-pass filtered signal; Step S5: Calculate the average peak interval of the found peak points and calculate the heart rate value based on the interval.

[0008] Preferably, step S1 comprises the following steps: Step S1.1: preprocess the data; Step S1.2: Perform wavelet transform on the preprocessed data and perform threshold processing on the wavelet coefficients using an adaptive threshold strategy; Step S1.3: Perform inverse transformation on the processed wavelet coefficients to obtain the processed signal.

[0009] Preferably, the adaptive threshold strategy is combined with a soft threshold function, which is dynamically adjusted according to the statistical characteristics of the signal and the noise level.

[0010] Preferably, step S2 comprises the following steps: Step S2.1: Perform Hilbert transform on the extracted heart rate signal to convert it into a complex signal; Step S2.2: Take the modulus of the complex signal, calculate the square root of the sum of the squares of its real part and imaginary part, and obtain the envelope of the signal.

[0011] Preferably, step S3 comprises the following steps: Step S3.1: Select the cutoff frequency and the filter order; Step S3.2: Apply the designed low-pass filter to the envelope signal to attenuate high-frequency noise.

[0012] Preferably, step S4 comprises the following steps: Step S4.1: Extend the signal forward and backward to provide comparison points for the data points at the boundary; Step S4.2: Use a multi-scale sliding window to compare the two sides of the signal and find the local maximum; Step S4.3: Determine the optimal window size by comparing the number of local maxima under different window sizes, and mark all peak points.

[0013] Preferably, the method further comprises performing difference calculation on the peak points found, subtracting the coordinate index of the previous peak point from the coordinate index corresponding to the next peak point to obtain an interval value, and calculating the heart rate value according to the interval value and the sampling rate.

[0014] The present invention also provides an adaptive multi-scale envelope peak BCG heart rate detection system, the system comprising the following modules: Module M1: collects BCG signal data and performs wavelet processing on the collected BCG signal data in combination with an adaptive threshold selection system; Module M2: Extract the envelope of the BCG signal data after wavelet processing; Module M3: low-pass filtering the extracted envelope; Module M4: Use the automatic multi-scale peak search system to search for peak points of the low-pass filtered signal; Module M5: Calculate the average peak interval of the peak points found and calculate the heart rate value based on the interval.

[0015] Preferably, the module M1 includes the following modules: Module M1.1: Preprocess the data; Module M1.2: Perform wavelet transform on the preprocessed data and threshold the wavelet coefficients using an adaptive threshold strategy; Module M1.3: Perform inverse transformation on the processed wavelet coefficients to obtain the processed signal; The adaptive threshold strategy combines a soft threshold function and dynamically adjusts according to the statistical characteristics of the signal and the noise level; The module M2 includes the following modules: Module M2.1: Perform Hilbert transform on the extracted heart rate signal and convert it into a complex signal; Module M2.2: Take the modulus of the complex signal, calculate the square root of the sum of the squares of its real and imaginary parts, and obtain the envelope of the signal; The module M3 includes the following modules: Module M3.1: Select the cutoff frequency and filter order; Module M3.2: Apply the designed low-pass filter to the envelope signal to attenuate high-frequency noise; The module M4 includes the following modules: Module M4.1: Expand the signal forward and backward to provide comparison points for data points at the boundary; Module M4.2: Use a multi-scale sliding window to compare the two sides of the signal and find the local maximum; Module M4.3: Determine the optimal window size by comparing the number of local maxima under different window sizes and mark all peak points.

[0016] Preferably, the system further comprises performing difference calculation on the peak points found, subtracting the coordinate index of the previous peak point from the coordinate index corresponding to the next peak point to obtain an interval value, and calculating the heart rate value according to the interval value and the sampling rate.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. By combining adaptive threshold and wavelet transform technology, the present invention can extract heart rate signals more effectively, avoiding the J peak identification instability problem that may occur in the traditional fixed threshold method on complex and changeable BCG signals; using envelope extraction and adaptive peak detection technology, the accuracy of heart rate monitoring is further enhanced, and the interference of noise and motion artifacts is reduced; 2. The adoption of the adaptive threshold selection method enables the present invention to be dynamically adjusted according to the statistical characteristics and noise levels of different signals, thereby improving the flexibility of signal processing; the application of the automatic multi-scale peak search method enables the present invention to flexibly identify the peaks in the heart rate signal without manually adjusting the parameters, thereby further enhancing the adaptability of signal processing; 3. Compared with the existing methods that require complex parameter systems and rely on heartbeat templates, the present invention achieves accurate extraction of BCG signals and heart rate calculation through a relatively simple technical solution, reduces the hardware resource requirements for embedded instruments and equipment, and is conducive to promotion and use in practical applications; 4. Before performing peak detection, the present invention ensures that the peak point omission can be effectively avoided without increasing the sampling frequency by expanding the signal forward and backward, thereby improving the accuracy of the calculation result; this is of great significance to the accuracy and reliability of heart rate monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 A schematic diagram of the present invention performing wavelet transformation on pre-processed data; Figure 2 Schematic diagram of waveform comparison before and after wavelet transformation of the present invention; Figure 3 It is a schematic diagram of the difference between the original data and the reconstructed data of the present invention; Figure 4 A schematic diagram of extracting an envelope of an extracted heart rate signal according to the present invention; Figure 5 A schematic diagram of the high-frequency noise and fast-changing parts of the envelope initially extracted by the present invention; Figure 6A schematic diagram of low-pass filtering of a signal after Hilbert transformation according to the present invention; Figure 7 This is a schematic diagram of the front and back expansion of a signal according to the present invention; Figure 8 A schematic diagram of reducing missed detections according to the present invention; Fig. 9 This is a schematic diagram of the present invention performing peak point detection on the expanded signal; Fig.10 According to the present invention, the coordinate index corresponding to the next peak point minus the coordinate index of the previous peak point is a schematic diagram of the interval value; Fig.11 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0019] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0020] Embodiment 1: Reference Fig.11 According to an adaptive multi-scale envelope peak BCG heart rate detection method provided by the present invention, the method comprises the following steps: Step S1: Collect BCG signal data, and perform wavelet processing on the collected BCG signal data in combination with the adaptive threshold selection method; Step S1.1: Preprocess the data; Step S1.2: Perform wavelet transform on the preprocessed data and use an adaptive threshold strategy to perform threshold processing on the wavelet coefficients; the adaptive threshold strategy combines a soft threshold function and performs dynamic adjustment according to the statistical characteristics of the signal and the noise level.

[0021] Step S1.3: Perform inverse transformation on the processed wavelet coefficients to obtain the processed signal.

[0022] Step S2: extracting the envelope of the BCG signal data after wavelet processing; Step S2.1: Perform Hilbert transform on the extracted heart rate signal to convert it into a complex signal; Step S2.2: Take the modulus of the complex signal, calculate the square root of the sum of the squares of its real part and imaginary part, and obtain the envelope of the signal.

[0023] Step S3: performing low-pass filtering on the extracted envelope; Step S3.1: Select the cutoff frequency and the order of the filter; the cutoff frequency is the frequency at which the attenuation of the signal by the filter begins to increase significantly after the signal frequency reaches this frequency when the signal passes through a filter, so that the signal's ability to pass through the filter becomes weaker. The main frequency components of the heart rate signal are usually between 0.5 and 40 Hz, which is the typical frequency range of human heart rate changes. Therefore, the cutoff frequency of the low-pass filter should be set within this range to retain the main information of the heart rate signal. If the noise is mainly concentrated in the high-frequency area, the cutoff frequency should be selected to remove these high-frequency noises while not affecting the low-frequency components of the heart rate signal. For example, common power frequency interference (such as 50Hz or 60Hz) needs to be filtered out, that is, the selected cutoff frequency should be lower than these interference frequencies. In addition, in practical applications, it is also necessary to determine the impact of the filter on the signal at different cutoff frequencies through experiments and select a suitable balance point. The order of the filter is related to the cutoff frequencies of the passband and stopband. The passband cutoff frequency and the stopband cutoff frequency determine the frequency range in which the filter needs to allow the signal to pass and start attenuating the signal. The higher the order of the filter, the better it can meet the attenuation requirements. However, the higher the filter order, the higher the computational complexity and implementation cost. Therefore, under the premise of meeting the filtering performance requirements, the lowest possible order should be selected to reduce resource consumption.

[0024] Step S3.2: Apply the designed low-pass filter to the envelope signal to attenuate high-frequency noise.

[0025] Step S4: using an automatic multi-scale peak search method to search for peak points of the low-pass filtered signal; Step S4.1: Extend the signal forward and backward to provide comparison points for the data points at the boundary; Step S4.2: Use a multi-scale sliding window to compare the two sides of the signal and find the local maximum; Step S4.3: Determine the optimal window size by comparing the number of local maxima under different window sizes, and mark all peak points.

[0026] Step S5: Find the average peak interval of the peak points found, and calculate the heart rate value based on the interval. The adaptive peak search algorithm can detect the J wave peak point in the BCG signal, which is the main feature point of the heart rate algorithm. By finding the position of two consecutive peak points and calculating the time difference between them, the time interval between adjacent peak points can be calculated (that is, the coordinate index corresponding to the latter peak point minus the coordinate index of the previous peak point to get the interval value), that is, the heartbeat interval. The heart rate value can be derived by calculating the average value of the heartbeat interval and converting it into the number of heartbeats per minute. The relationship between the interval and the heart rate can be calculated by the formula: Heart rate = ((60*sampling rate) / average heartbeat interval) times / minute The average heartbeat interval is the average of the intervals between consecutive peaks. Specifically, if n peaks are detected from the BCG signal, there will be n-1 intervals between these peaks. To calculate the average heartbeat interval, the sum of these intervals needs to be divided by the number of intervals (n-1). The average value obtained in this way is the average heartbeat interval. That is: Mean = sum of all interval values ​​ / n-1.

[0027] The present invention also provides an adaptive multi-scale envelope peak point BCG heart rate detection system, which can be implemented by executing the process steps of the adaptive multi-scale envelope peak point BCG heart rate detection method, that is, those skilled in the art can understand the adaptive multi-scale envelope peak point BCG heart rate detection method as a preferred implementation of the adaptive multi-scale envelope peak point BCG heart rate detection system.

[0028] Embodiment 2: The present invention also provides an adaptive multi-scale envelope peak BCG heart rate detection system, the system comprising the following modules: Module M1: collects BCG signal data and performs wavelet processing on the collected BCG signal data in combination with an adaptive threshold selection system; Module M1.1: Preprocess the data; Module M1.2: Perform wavelet transform on the preprocessed data, and use adaptive threshold strategy to threshold the wavelet coefficients; the adaptive threshold strategy combines the soft threshold function and dynamically adjusts according to the statistical characteristics of the signal and the noise level. The threshold processing of wavelet transform mainly includes two parts, one is the selection of threshold, and the other is the selection of threshold function. The selection of threshold is used to judge the noise, and its size determines the quality of denoising. The selection of adaptive threshold means that the threshold can be adaptively selected according to the noise level. For example, the VisuShrink threshold (universal threshold) selection method, this threshold is obtained for the joint distribution of multi-dimensional independent normal variables under the Gaussian model. This method takes into account the local characteristics of the signal and the standard deviation of the noise, so that the threshold can be dynamically adjusted to adapt to different signal and noise conditions. The threshold function is used to perform threshold shrinkage processing on the wavelet coefficient ω. The threshold function is generally divided into hard threshold and soft threshold. Here, the soft threshold function is selected as a further optimization of the adaptive threshold selection, that is, the size of the threshold is further adjusted according to the energy distribution of the wavelet coefficient. In summary, the adaptive threshold strategy used here is a combination of adaptive threshold selection and soft threshold function, that is, by analyzing the mean and variance of the signal, the local characteristics of the signal are determined, and the dynamic adjustment of the threshold is guided by these characteristics.

[0029] Module M1.3: Perform inverse transformation on the processed wavelet coefficients to obtain the processed signal; Module M2: Extract the envelope of the BCG signal data after wavelet processing; Module M2.1: Perform Hilbert transform on the extracted heart rate signal and convert it into a complex signal; Module M2.2: Take the modulus of the complex signal, calculate the square root of the sum of the squares of its real and imaginary parts, and obtain the envelope of the signal; Module M3: low-pass filtering the extracted envelope; Module M3.1: Select the cutoff frequency and filter order; Module M3.2: Apply the designed low-pass filter to the envelope signal to attenuate high-frequency noise; Module M4: Use the automatic multi-scale peak search system to search for peak points of the low-pass filtered signal; Module M4.1: Expand the signal forward and backward to provide comparison points for data points at the boundary; Module M4.2: Use a multi-scale sliding window to compare the two sides of the signal and find the local maximum; Module M4.3: Determine the optimal window size by comparing the number of local maxima under different window sizes and mark all peak points.

[0030] Module M5: Calculate the average peak interval of the peak points found and calculate the heart rate value based on the interval.

[0031] The system also includes performing difference calculation on the peak points found, subtracting the coordinate index of the previous peak point from the coordinate index corresponding to the next peak point to obtain an interval value, and calculating the heart rate value according to the interval value and the sampling rate.

[0032] Embodiment 3: The present invention extracts the heart rate signal by using adaptive threshold and wavelet transform technology, combines envelope extraction and adaptive peak detection technology to enhance the accuracy of heart rate monitoring, and includes the following steps: Step 1: Combine the adaptive threshold selection method to perform wavelet processing on the data; Step 2: Extract the envelope of the wavelet processed data; Step 3: low-pass filter the extracted envelope; Step 4: Use the automatic multi-scale peak search method to search for peak points of the low-pass filtered signal.

[0033] Step 5: Find the average peak interval of the peak points found, and calculate the heart rate value based on the interval.

[0034] The step 1 comprises the following steps: Step 1.1: Perform data preprocessing on the collected data to reduce the impact of impulse noise on signal processing. At this stage, a sliding average filter is added to smooth the data. The average value of each point is calculated by selecting an appropriate window size, and the calculated average value is assigned to the corresponding position of the output signal until the entire signal is processed. The choice of window size depends on the characteristics of the signal and the required smoothness. A larger window will produce smoother results, but may also introduce more delays.

[0035] Step 1.2: Preprocess the data Figure 1 Wavelet transform is performed. Wavelet transform can decompose the signal into sub-bands of different frequencies and process it using the sparsity of wavelet coefficients. For BCG signals, selecting appropriate wavelet basis functions and decomposition layers can obtain better signal representation. Here, taking the db5 wavelet basis as an example, the preprocessed signal is subjected to 3-layer wavelet decomposition, and the wavelet coefficients are threshold processed using an adaptive threshold strategy. The adaptive threshold strategy can be dynamically adjusted according to the statistical characteristics and noise level of the signal to achieve better processing results. Common threshold selection strategies include general threshold, maximum and minimum threshold, and mixed threshold. The appropriate threshold selection strategy can be selected according to the specific application scenario.

[0036] Step 1.3: According to the selected threshold strategy, threshold processing is performed on the wavelet coefficients. For coefficients less than the threshold, they are reduced according to the soft threshold strategy. Then the processed wavelet coefficients are inversely transformed to obtain the processed signal. Figure 2 The waveform comparison before and after wavelet transformation is given in . The final required heart rate signal is a high-frequency signal, that is, the difference between the original data and the reconstructed data Figure 3 .

[0037] The step 2 comprises the following steps: Step 2.1: Extract the envelope of the extracted heart rate signal Figure 4 . The envelope of the signal reveals the changing trend and overall characteristics of the signal over a long time scale. The real signal is converted into a complex signal by using the Hilbert transform, and the signal is analyzed by using the filter in the imaginary circuit, where the amplitude spectrum of the analytical signal represents the envelope information of the original signal, that is, the analytical signal is modulo, and the square root of the sum of the squares of its real part and imaginary part is calculated to obtain the envelope of the original signal. In this way, key features in the BCG signal, such as heartbeat cycle and amplitude changes, can be more clearly identified. However, the initially extracted envelope may contain high-frequency noise and rapidly changing parts. Figure 5 , which will affect the accurate identification of the signal peak point.

[0038] The step 3 comprises the following steps: Step 3.1: Low-pass filter the Hilbert transformed signal Figure 6 . Select the appropriate cutoff frequency and filter order, and apply the designed low-pass filter to the envelope signal. The low-pass filter allows low-frequency signals to pass through while attenuating high-frequency signals. Since the high-frequency components of the noise are attenuated by the filter, the main features of the signal (usually at lower frequencies) are retained, so the filtered signal will be cleaner and reduce noise interference.

[0039] The step 4 comprises the following steps: Step 4.1: Extend the signal forward and backward Figure 7 In the BCG signal, the peak point usually corresponds to a specific phase of the heartbeat, such as the J point, which has the most obvious amplitude characteristics and is consistent with the heartbeat cycle. By detecting these peak points, the timing information of the heartbeat can be determined. Therefore, in order to ensure that the automatic multi-scale peak finding method has the same number of data points for comparison at each data point of the signal (including the first and the last), the first Figure 8 By extending the signal forward and backward, it is possible to provide "virtual" comparison points for the data points at the boundaries. These values ​​do not affect the overall characteristics of the signal, but provide the algorithm with the necessary data points to enable it to detect peaks at the beginning and end of the signal.

[0040] Step 4.2: Use the automatic multi-scale peak finding method to detect the peak point of the expanded signal Fig. 9 . The algorithm uses a multi-scale sliding window to compare both sides of the signal to find the local maximum. The core of the algorithm is to use different window sizes (k values) to detect peaks and find the optimal window size. That is, for each possible window size k, from k to the signal length minus k, check whether each point is greater than the value of the k points before and after. If so, the point is considered to be a local maximum and the local maxima are counted. By comparing the number of local maxima under different window sizes, find the window size that makes the number of local maxima the largest. This window size is the maximum window length. Using the determined maximum window length, traverse the signal again and mark all points that meet the local maximum conditions. These points are peak points.

[0041] The step 5 comprises the following steps: Step 5.1: Calculate the difference of the peak points found. Subtract the coordinate index of the previous peak point from the coordinate index of the next peak point to get the interval value. Fig.10 The relationship between interval and heart rate can be calculated by the formula: heart rate = ((60*sampling rate) / average heartbeat interval) times / minute.

[0042] The principle is based on the following points: 1) Heart rate is the number of times the heart beats per minute. It is inversely proportional to the heartbeat interval (i.e., the time between two heartbeats). If the heartbeat interval is T seconds, then the heart rate is 1 / T times / second; 2) Since the heart rate is usually measured in beats / minute and the heartbeat interval is measured in seconds, seconds need to be converted into minutes, so the calculation formula for heart rate includes the coefficient 60; 3) Sampling rate refers to the number of times a signal is collected per second, which is related to the discrete time points of the signal. In digital signal processing, the sampling rate determines the degree of discretization of time. Therefore, in order to convert discrete time points into continuous time intervals, the sampling rate needs to be taken into account; 4) The average heartbeat interval refers to the average of all heartbeat intervals within a certain period of time. Since the heartbeat interval may fluctuate due to various factors, taking the average value can more accurately reflect the heart rate.

[0043] To sum up, taking into account various factors, the complete calculation process is as follows: First, convert the average heartbeat interval from discrete time points to continuous time intervals, that is, divide it by the sampling rate; second, convert this continuous time interval into minutes, that is, divide it by 60; finally, take the reciprocal to get the heart rate, that is, the number of heartbeats per minute.

[0044] The present invention adopts an adaptive threshold selection method to effectively solve the deficiency of using the same threshold at each level of scale. The combination with the soft threshold function can better retain the basic characteristics of the signal, avoid the signal distortion that may be caused by hard threshold processing, and help improve the quality of signal reconstruction. Through envelope extraction, the key features in the heart rate signal can be more clearly identified, and the high-frequency noise and other interference in the signal can be effectively removed, thereby improving the accuracy of signal processing. The use of low-pass filtering can reduce the influence of artifacts, making the heart rate signal clearer, helping to more accurately detect the peak value of the heartbeat and reduce the heart rate estimation error. Before performing peak detection, the front and back expansion of the signal ensures that the peak point can be effectively avoided without increasing the sampling frequency, thereby improving the accuracy of the calculation result. By adopting an automatic multi-scale peak search method, the required peak value can be flexibly identified in the heart rate signal without manually adjusting the parameters, thereby improving the adaptability of the heart rate signal peak detection.

[0045] Those skilled in the art may understand this embodiment as a more specific description of Embodiment 1 and Embodiment 2.

[0046] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0047] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. An adaptive multi-scale envelope peak BCG heart rate detection method, characterized in that: The method comprises the following steps: Step S1: Collect BCG signal data, and perform wavelet processing on the collected BCG signal data in combination with an adaptive threshold selection method; Step S2: extracting the envelope of the BCG signal data after wavelet processing; Step S3: performing low-pass filtering on the extracted envelope; Step S4: using an automatic multi-scale peak search method to search for peak points of the low-pass filtered signal; Step S5: Calculate the average peak interval of the peak points found, and calculate the heart rate value according to the interval; The step S1 comprises the following steps: Step S1.1: preprocess the data; Step S1.2: Perform wavelet transform on the preprocessed data and perform threshold processing on the wavelet coefficients using an adaptive threshold strategy; Step S1.3: Perform inverse transformation on the processed wavelet coefficients to obtain the processed signal; The adaptive threshold strategy combines a soft threshold function and makes dynamic adjustments based on the statistical characteristics of the signal and the noise level.

2. The adaptive multi-scale envelope peak BCG heart rate detection method according to claim 1, characterized in that: The step S2 comprises the following steps: Step S2.1: Perform Hilbert transform on the extracted heart rate signal to convert it into a complex signal; Step S2.2: Take the modulus of the complex signal, calculate the square root of the sum of the squares of its real part and imaginary part, and obtain the envelope of the signal.

3. The adaptive multi-scale envelope peak BCG heart rate detection method according to claim 1, characterized in that: The step S3 comprises the following steps: Step S3.1: Select the cutoff frequency and the filter order; Step S3.2: Apply the designed low-pass filter to the envelope signal to attenuate high-frequency noise.

4. The adaptive multi-scale envelope peak BCG heart rate detection method according to claim 1, characterized in that: The step S4 comprises the following steps: Step S4.1: Extend the signal forward and backward to provide comparison points for the data points at the boundary; Step S4.2: Use a multi-scale sliding window to compare the two sides of the signal and find the local maximum; Step S4.3: Determine the optimal window size by comparing the number of local maxima under different window sizes, and mark all peak points.

5. The adaptive multi-scale envelope peak BCG heart rate detection method according to claim 1, characterized in that: The method further includes performing difference calculation on the peak points found, subtracting the coordinate index of the previous peak point from the coordinate index corresponding to the next peak point to obtain an interval value, and calculating the heart rate value according to the interval value and the sampling rate.

6. An adaptive multi-scale envelope peak BCG heart rate detection system, characterized in that: The system includes the following modules: Module M1: collects BCG signal data and performs wavelet processing on the collected BCG signal data in combination with an adaptive threshold selection system; Module M2: Extract the envelope of the BCG signal data after wavelet processing; Module M3: low-pass filtering the extracted envelope; Module M4: Use the automatic multi-scale peak search system to search for peak points of the low-pass filtered signal; Module M5: Calculate the average peak interval of the peak points found, and calculate the heart rate value based on the interval; The module M1 includes the following modules: Module M1.1: Preprocess the data; Module M1.2: Perform wavelet transform on the preprocessed data and threshold the wavelet coefficients using an adaptive threshold strategy; Module M1.3: Perform inverse transformation on the processed wavelet coefficients to obtain the processed signal; The adaptive threshold strategy combines a soft threshold function and makes dynamic adjustments based on the statistical characteristics of the signal and the noise level.

7. The adaptive multi-scale envelope peak BCG heart rate detection system according to claim 6, characterized in that: The module M2 includes the following modules: Module M2.1: Perform Hilbert transform on the extracted heart rate signal and convert it into a complex signal; Module M2.2: Take the modulus of the complex signal, calculate the square root of the sum of the squares of its real and imaginary parts, and obtain the envelope of the signal; The module M3 includes the following modules: Module M3.1: Select the cutoff frequency and filter order; Module M3.2: Apply the designed low-pass filter to the envelope signal to attenuate high-frequency noise; The module M4 includes the following modules: Module M4.1: Expand the signal forward and backward to provide comparison points for data points at the boundary; Module M4.2: Use a multi-scale sliding window to compare the two sides of the signal and find the local maximum; Module M4.3: Determine the optimal window size by comparing the number of local maxima under different window sizes and mark all peak points.

8. The adaptive multi-scale envelope peak BCG heart rate detection system according to claim 6, characterized in that: The system also includes performing difference calculation on the peak points found, subtracting the coordinate index of the previous peak point from the coordinate index corresponding to the next peak point to obtain an interval value, and calculating the heart rate value according to the interval value and the sampling rate.

Images