Apparatus and method for extracting heart rate data based on wireless radar signals

Abstract

To provide a device and a method for extracting heart beat data on the basis of a radio radar signal.SOLUTION: A method includes: collecting a radio radar signal obtained by sensing a detection target using radar; performing band-pass filtering in a first frequency band on radio radar data; calculating frequency spectrum energy at each time accumulated in a time window; acquiring band-pass filtering feature data by performing feature extraction on data after the band-pass filtering and / or acquiring frequency spectrum energy feature data by performing feature extraction on frequency spectrum energy at each time in a second frequency band; inputting the band-pass filtering feature data and / or the frequency spectrum energy feature data to a neural network based detection model; and obtaining a heart beat interval and the number of heart beats of the detection target on the basis of waveform data.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to the technical field of radar detection.

Background Art

[0002] With the improvement of living standards, people's health awareness has been gradually increasing, and it is necessary to regularly evaluate the heartbeat and monitor the health status. Generally speaking, there are two indicators for evaluating the heartbeat. One is the resting heart rate, and the other is the heart rate variability (HRV). The resting heart rate refers to the number of heartbeats per minute of a normal person in a resting state, and the heart rate variability refers to the change in the difference between heartbeat cycles. The indicators for evaluating the heartbeat are very important guidelines for disease diagnosis and health management, and the heart rate variability can also be used for evaluating the stress index, identifying emotions, etc.

[0003] In medicine, usually, an electrocardiogram (ECG) is used to monitor the patient's resting heart rate and heart rate variability. The monitoring by electrocardiogram is very accurate, but it is necessary to stick a sheet on the designated position of the skin during the test, so it is not user-friendly, there is a sense of discomfort, and it is inconvenient to perform monitoring using the electrocardiogram at any time.

[0004] With the popularization of smart wearable devices, heart rate monitoring has been integrated into most smart bracelets / watches. However, most of these devices can only provide the heart rate index, and it is difficult to accurately evaluate the heartbeat. Moreover, because it is necessary to wear them for a long time, the elderly are likely to forget and it is troublesome to wear. On the other hand, at present, non-contact monitoring of the heartbeat has been proposed. For example, by analyzing the radar signal reflected by the human body, a signal related to the heartbeat is extracted.

[0005] The above-mentioned introduction of background art is intended to clearly and completely explain the proposed technical aspects of the present invention and to facilitate understanding by those skilled in the art. These technical aspects, as described in the background art of the present invention, should not be construed as being well-known to those skilled in the art. [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] However, the inventors discovered the following: Since physiological signals include not only heartbeat signals but also respiratory signals, the extracted heartbeat signals are susceptible to influence from respiratory signals. Currently, non-contact heartbeat monitoring schemes can generally detect the presence or absence of a heartbeat signal, but the detection accuracy is low, and there is a risk that accurate heartbeat data cannot be extracted.

[0007] In view of at least one of the technical problems described above, embodiments of the present invention provide an apparatus and method for extracting heartbeat data based on a wireless radar signal. Accurate, non-contact acquisition of heartbeat data enables monitoring of resting heart rate and heart rate variability. [Means for solving the problem]

[0008] According to one aspect of an embodiment of the present invention, an apparatus is provided for extracting heartbeat data based on a radio radar signal, the apparatus is A collection unit for collecting radio radar signals detected by radar; A filtering unit for performing bandpass filtering of a first frequency band on an acquired radio radar signal, wherein the first frequency band is a predetermined frequency band related to the heartbeat of the object to be detected; A computing unit for calculating the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; A feature extraction unit for obtaining bandpass filtering feature data by performing feature extraction on the data after bandpass filtering, and / or obtaining frequency spectral energy feature data by performing feature extraction on the frequency spectral energy at each time point within the second frequency band; A detection unit for inputting the time-domain bandpass filtering feature data and / or the frequency-domain frequency spectral energy feature data into a neural network-based detection model and obtaining the detected waveform data using the detection model; and The unit includes an acquisition unit for obtaining the heartbeat interval and heart rate of the target to be detected based on the waveform data.

[0009] According to another aspect of the embodiments of the present invention, a method is provided for extracting heartbeat data based on a radio radar signal, the method being: The radar collects radio radar signals that detect an object; The acquired radio radar signal is subjected to bandpass filtering of a first frequency band, the first frequency band being a predetermined frequency band related to the heartbeat of the target being detected; Calculate the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; Feature extraction is performed on the data after the bandpass filtering to obtain bandpass filtering feature data, and / or feature extraction is performed on the frequency spectral energy at each time point within the second frequency band to obtain frequency spectral energy feature data; The time-domain bandpass filtering feature data and / or the frequency-domain frequency spectral energy feature data are input to a neural network-based detection model, and the detected waveform data is obtained using the detection model; and This includes obtaining the heartbeat interval and heart rate of the target to be detected based on the waveform data. [Effects of the Invention]

[0010] The advantageous effects of the embodiments of the present invention are at least as follows: Bandpass filtering feature data and / or frequency spectral energy feature data are acquired based on a wireless radar signal; the time-domain bandpass filtering feature data and / or frequency-domain frequency spectral energy feature data are input into a neural network-based detection model; waveform data after detection is acquired using the detection model; and the heartbeat interval and heart rate of the target of detection can be acquired based on the waveform data. As a result, heartbeat data can be acquired accurately and non-contact, enabling monitoring of resting heart rate and heart rate variability. This makes heartbeat monitoring convenient at any time, provides a comfortable user experience, and offers high accuracy.

[0011] Furthermore, features described and / or shown in one embodiment may be used in the same or similar manner in one or more other embodiments, combined with or substituting features in other embodiments.

[0012] When used herein, terms such as “contains / have” refer to the presence of a feature, element, step, or assembly, but do not exclude the presence or addition of one or more other features, elements, steps, or assemblies. [Brief explanation of the drawing]

[0013] Elements and features described in one drawing or one embodiment of the present invention can be combined with elements and features shown in one or more other drawings or embodiments. Furthermore, similar reference numerals in the drawings are used to indicate corresponding parts in several drawings and to indicate corresponding parts used in multiple embodiments. [Figure 1] This figure shows the electrocardiogram (ECG) signal obtained by an electrocardiogram (ECG). [Figure 2] This figure shows a method for extracting heartbeat data based on a wireless radar signal in an embodiment of the present invention. [Figure 3]It is a diagram showing an example of wireless radar data in an embodiment of the present invention. [Figure 4] It is a diagram showing an example of data after band-pass filtering in an embodiment of the present invention. [Figure 5] It is a diagram showing an example of frequency spectrum energy at a plurality of times in an embodiment of the present invention. [Figure 6] It is a diagram showing an example of an electrocardiogram signal acquired by an ECG in an embodiment of the present invention. [Figure 7] It is a diagram showing an example of band-pass filtering characteristic data in an embodiment of the present invention. [Figure 8] It is a diagram showing another method of extracting heartbeat data based on a wireless radar signal in an embodiment of the present invention. [Figure 9] It is a diagram showing an example of comparison between an electrocardiogram signal acquired by an ECG and triangular wave data in an embodiment of the present invention. [Figure 10] It is a diagram showing an example of loss and histogram of heart rate and heartbeat interval in an embodiment of the present invention. [Figure 11] It is a diagram showing an apparatus for extracting heartbeat data based on a wireless radar signal in an embodiment of the present invention [Figure 12] It is a diagram showing an electronic device in an embodiment of the present invention.

Mode for Carrying Out the Invention

[0014] By referring to the accompanying drawings and the following description, the foregoing and other features of the present invention will become apparent. In the specification and drawings, specific embodiments of the present invention are disclosed, but they show only some of the embodiments that can adopt the principles of the present invention. It should be understood that the present invention is not limited to the described embodiments, that is, the present invention includes all changes, modifications and substitutions belonging to the appended claims.

[0015] In embodiments of the present invention, the radar may be, but is not limited to, a millimeter-wave (mmWave) radar. The radar transmits electromagnetic waves using a transmitting antenna and receives corresponding reflected waves (radar echo information) after they have been reflected by different objects. By analyzing the radar echo information, information such as the position between the object and the radar and its radial velocity can be effectively extracted, and this information can meet the needs of many application scenarios.

[0016] In embodiments of the present invention, the object to be detected may be people of various ages, for example, elderly people, children, elderly people and / or caregivers, children and / or guardians. However, the present invention is not limited to these, and the object to be detected may also be an animal with biological characteristics. The following explanation will use the human body as an example.

[0017] Figure 1 shows an electrocardiogram (ECG) signal obtained by an electrocardiogram. As shown in Figure 1, the ECG signal has at least two R-wave information (wave peaks) 101. The inter-beat interval (IBI) and heart rate can be calculated from this ECG signal.

[0018] In embodiments of the present invention, the heartbeat data includes the heartbeat interval to be detected (which may be expressed using IBI) and the heart rate, the heart rate can represent an index called resting heart rate, and the heartbeat interval can be used to analyze heart rate variability. For specific details of these concepts, please refer to related technologies.

[0019] <Example of the first side view> An embodiment of the present invention provides a method for extracting heartbeat data based on a wireless radar signal.

[0020] Figure 2 shows a method for extracting heartbeat data based on a wireless radar signal in an embodiment of the present invention, and as shown in Figure 2, the method includes the following operations (steps).

[0021] 201: Collect the radio radar signal that detects the target using radar; 202: Bandpass filtering of a first frequency band is performed on the acquired (collected) wireless radar data, wherein the first frequency band is a predetermined frequency band related to the heartbeat of the detected object; 203: Calculate the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; 204: Perform feature extraction on the data after bandpass filtering to obtain bandpass filtering feature data, and / or perform feature extraction on the frequency spectral energy at each time point within the second frequency band to obtain frequency spectral energy feature data; 205: Input time-domain bandpass filtering feature data and / or frequency-domain frequency spectral energy feature data into a neural network-based detection model, and obtain the detected waveform data using the detection model; and 206: Based on the waveform data, the heartbeat interval and heart rate of the target to be detected are obtained.

[0022] Figure 2 above is merely illustrative to illustrate an embodiment of the present invention, and the present invention is not limited thereto. For example, the execution order between each operation (step) can be appropriately adjusted, or some operations can be added or removed. Furthermore, those skilled in the art can make appropriate modifications based on the above description, not limited to the description in Figure 2.

[0023] In some embodiments, a single-frequency modulated continuous-wave radar system is employed, and the external space can be sensed by a radio signal transmitted by the radar transmitter. The object to be detected is within the radiation range of the radar system, and can, for example, be sitting or lying down at a specific distance from the radar system, for example, less than 60 cm away. A single radar system may monitor a certain object to be detected, but the present invention is not limited thereto, and for example, an enhanced radar system can monitor multiple objects to be detected simultaneously.

[0024] In some embodiments, a radar receiver can receive radar echoes (wireless radar signals), and a data acquisition card can collect these wireless radar signals at a first sampling rate S1 to obtain wireless radar data. This first sampling rate S1 may be relatively high, for example, 2500 Hz.

[0025] Figure 3 shows an example of wireless radar data in an embodiment of the present invention, and as shown in Figure 3, the wireless radar data includes, for example, an in-phase signal radar_I(n) and an orthogonal signal radar_Q(n).

[0026] In some embodiments, the first frequency band is a predetermined frequency band relating to the heartbeat to be detected, for example, the first frequency band is [8,100] Hz or [-100,-8][8,100] Hz. A bandpass filter can provide a basis for accurate extraction of heartbeat data by eliminating to some extent the influence of low-frequency respiratory signals and high-frequency noise signals.

[0027] For example, by performing bandpass filtering on radar_I(n) and radar_Q(n), it is possible to obtain data that has been bandpass filtered and is primarily modulated by heartbeat waves, namely radar_I_bp(n) and radar_Q_bp(n).

[0028] In some embodiments, the DC frequency offset may be further removed from the data after bandpass filtering.

[0029] For example, the following operation, namely, radar_I_bp(n)=radar_I_bp(n)-mean(radar_I_bp(n)); and radar_Q_bp(n)=radar_Q_bp(n)-mean(radar_Q_bp(n)) This process can eliminate the DC offset.

[0030] Figure 4 shows an example of data after bandpass filtering in an embodiment of the present invention. As shown in Figure 4, the data after bandpass filtering includes, for example, the common-mode signal radar_I_bp(n) and the quadrature signal radar_Q_bp(n) after DC offset cancellation.

[0031] In some embodiments, the frequency spectral energy at each time point can be obtained by converting the bandpass filtered data from a time-domain signal to a frequency-domain signal using a sliding time window. For example, the bandpass filtered data can be converted from a time-domain signal to a frequency-domain signal using methods such as the Short-Time Fourier Transform (STFT) or wavelet transform.

[0032] For example, the processed IQ complex signal is converted from the time domain to the frequency domain, and the complex signal is, S(n)=radar_I_bp(n)+j*radar_Q_bp(n) The time window may also be represented by a sequence, where n represents each time step.

[0033] For example, a complex signal S(n) can be divided into n-stage (k-point) sequences S(k), a frequency domain transformation (e.g., FFT) can be performed on each sequence S(k,i) to obtain the frequency spectrum z(t(i),f(m),i) of that sequence, and after combining the frequency spectra of each sequence, the frequency spectrum z(t(n),f(m)) of the complex signal S(n) can be obtained.

[0034] In this sequence, S(k,i) represents the i-th sequence with sequence length k, where i = 0, 1, 2, ..., n, and k represents the number of points in the sequence where the frequency domain transformation is performed. The number of points may be a fixed value, such as 512. z(t(i),f(m),i) represents the i-th sequence (i.e., the time is t(i), and t(i) = i*f s and f sThis represents the frequency spectral energy at each frequency point f(m) (m=0,1,2,…,k-1) of the sampling frequency.

[0035] Figure 5 shows an example of the frequency spectral energy at multiple time points in an embodiment of the present invention. The segmented time window is, for example, 204.8 ms (512-point FFT), and the sliding step length is 0.4 ms. For example, by sliding point by point, it is possible to ensure that the sampling rate matches the sampling rate of the IQ signal. For example, the frequency spectral energy at each time point can be obtained within the range of [-30,-8] and [8,30] Hz (as shown by the grayish-white area in Figure 5).

[0036] Figure 6 shows an example of an electrocardiogram signal acquired by ECG in an embodiment of the present invention. As shown in Figures 5 and 6, the frequency spectral energy distribution in Figure 5 is in close agreement with the electrocardiogram signal distribution in Figure 6. This allows for initial verification that the embodiment of the present invention can improve the accuracy of heartbeat data.

[0037] First, we will explain how to obtain frequency spectral energy feature data.

[0038] In some embodiments, the second frequency band may be a subset of the first frequency band. For example, the second frequency band may be [8,30]Hz or [-30,-8][8,30]Hz. By performing frequency spectrum aggregation and analysis within the second frequency band, the influence of noise signals can be further removed, and the accuracy of heartbeat data can be further improved.

[0039] In some embodiments, frequency spectral energy feature data is obtained by performing feature extraction on the frequency spectral energy at each time point within the second frequency band, and the frequency spectral energy feature data includes one or more of the following, namely, This data is obtained after calculating the sum of the frequency spectral energies at each time point in a portion of the frequency band within the aforementioned second frequency band. For example, the data after calculating the sum for the frequency spectral energy in the frequency band [8,30] Hz, i.e.,

number

number

number

number

number

number

[0040] Although frequency spectral energy feature data has been described exemplarily above, the present invention is not limited thereto, and other feature data obtained after feature extraction based on frequency spectral energy may also be used. Some or all of this frequency spectral energy feature data can be input into a neural network-based detection model, and waveform data after detection can be obtained using the detection model.

[0041] Next, we will explain how to obtain bandpass filtering feature data.

[0042] In some embodiments, bandpass filtering feature data can also be obtained by performing feature extraction on the data after bandpass filtering. The bandpass filtering feature data may include one or more of the following: This data is obtained after applying bandpass filtering to the orthogonal signals of the aforementioned radar data. For example, radar_Q(n); This data is obtained after applying bandpass filtering to the in-phase signals of the aforementioned radar data. For example, radar_I_bp(n); This data is obtained after applying bandpass filtering to the aforementioned radar data to determine the amplitude values. for example,

number

[0043] Figure 7 shows an example of bandpass filtering feature data in an embodiment of the present invention, and shows the data radar_I_bp after bandpass filtering and removal of the DC offset, the data radar_Q_bp after bandpass filtering and removal of the DC offset, and the data iq_am after the amplitude value has been determined.

[0044] Although the bandpass filtering feature data has been described exemplarily above, the present invention is not limited thereto, and other feature data obtained after feature extraction based on bandpass filtering data may also be used. By inputting some or all of this bandpass filtering feature data into a neural network-based detection model and using the detection model to detect the frequency spectral energy feature data and the bandpass filtering feature data, the detected waveform data can be obtained.

[0045] Figure 8 illustrates another method for extracting heartbeat data based on a radio radar signal in an embodiment of the present invention, showing a case where both bandpass filtering feature data and frequency spectral energy feature data are used as feature data input to the detection model. As shown in Figure 8, the method includes the following steps (operations):

[0046] 801: Collects radio radar signals that detect an object using radar; 802: Bandpass filtering of a first frequency band is performed on the acquired radio radar data, wherein the first frequency band is a predetermined frequency band related to the heartbeat of the object to be detected; 803: Calculate the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; 804: Frequency spectral energy feature data is obtained by performing feature extraction on the frequency spectral energy at each time point within the second frequency band.

[0047] As shown in Figure 8, the method further includes the following steps.

[0048] 805: Bandpass filtering feature data is obtained by performing feature extraction on the data after bandpass filtering; 806: Input frequency spectral energy feature data and bandpass filtering feature data into a neural network-based detection model, and obtain the detected waveform data using the detection model; and 807: Based on the waveform data, the heartbeat interval and heart rate of the target to be detected are obtained.

[0049] It should be noted that Figure 8 above is merely illustrative to illustrate an embodiment of the present invention, and the present invention is not limited thereto. For example, the execution order between each operation can be appropriately adjusted, or some operations can be added or removed. Furthermore, those skilled in the art can make appropriate modifications based on the above description, not limited to the description in Figure 8.

[0050] In some embodiments, before inputting the feature data (frequency spectral energy feature data and / or bandpass filtering feature data) into a neural network-based detection model, the feature data may be further resampled at a second sampling rate S2, where the second sampling rate S2 is lower than the first sampling rate S1.

[0051] For example, the first sampling rate S1 is 2500 Hz and the second sampling rate S2 is 200 Hz. The relatively high first sampling rate allows for the acquisition of radio radar signals that accurately reflect the characteristics of the heartbeat, thus obtaining accurate feature data. The relatively low second sampling rate helps to prevent data distortion and reduces the complexity of the deep learning model.

[0052] In some embodiments, normalization may be performed on the feature data after resampling.

[0053] For example, the radar feature information is normalized using the min-max normalization method. Taking radar_I_bp as an example,

number

[0054] In some embodiments, feature data within a single time period may be input to the detection model, where this single time period is a predetermined time (e.g., 2 seconds) that ensures it includes at least two R-wave information points. This further guarantees the accuracy of the detection data and provides a fast response time.

[0055] In some embodiments, the waveform data of the detection model output is triangular wave data, which allows for easy processing of the waveform data.

[0056] Figure 9 shows an example of a comparison between an electrocardiogram signal acquired by ECG and triangular wave data in an embodiment of the present invention. As shown in Figure 9, the loss (shown by the polyline in Figure 9) between the triangular wave data and the electrocardiogram signal acquired by ECG in the embodiment of the present invention is controlled to stay within a certain range, thus the accuracy of the heartbeat data in the embodiment of the present invention is high.

[0057] In some embodiments, noise may be added to the triangular wave data output from the detection model; and the maximum value of the triangular wave data after noise addition may be detected to obtain the heartbeat interval and heart rate of the target to be detected.

[0058] For example, to avoid detecting the same local maximum, ypred=ypred+rand(length(ypred))*0.0001 As shown above, a small amount of noise can be added to triangular wave data.

[0059] In some embodiments, the heartbeat interval and heart rate of the target can be obtained based on the triangular wave data after noise has been added. Specifically, the maximum value of the triangular wave data is associated with the R-wave information of the electrocardiogram signal obtained by the electrocardiogram, the heartbeat interval of the target can be obtained based on the time interval of adjacent maximum values, and the heart rate of the target can be obtained based on the obtained heartbeat interval of the target. This allows for the detection of the maximum value of the triangular wave after noise has been added, and by associating the position of the maximum value point with the position of the R-wave in the electrocardiogram, the accuracy of detection can be further improved.

[0060] For example, the IBI at the current time can be obtained from the values ​​of adjacent R-wave intervals (RR intervals) within a window with an average of 5 seconds. The data sampling rate is s0, the window time length is 5 seconds, there are m R-waves, and the corresponding positions of the R-waves are {l0, l1, ..., l m-1 If it is},

number

number

[0061] Figure 10 shows an example of heart rate and heartbeat interval loss and histogram in an embodiment of the present invention. As shown in Figure 10, the scheme of the embodiment of the present invention allows for accurate extraction of heartbeat data from a wireless radar signal.

[0062] The above provides an illustrative explanation of how to extract heartbeat data; below, we will briefly explain the detection model and training.

[0063] In some embodiments, the neural network-based detection model may be a deep learning neural network model, such as ResNet or VGG, or a 1D fully convolutional network, and the loss function may be KLDivLoss, etc. During the training process, the electrocardiogram signals detected by the ECG may be used as true values ​​for training.

[0064] In the embodiments of the present invention, one or more sets of optimal parameters may be obtained by supervised training (learning), and then the parameters are applied to the detection model. In the embodiments of the present invention, the structure of the detection model is not limited, and related technologies can be referenced. Furthermore, the specific training method is not limited, for example, SGD (Stochastic Gradient Descent) optimization, Adam (Adaptive) optimization, etc. You may also use optimizations such as Moment Estimation.

[0065] Although only the steps or processes of the present invention have been described above, the present invention is not limited thereto. The method for extracting heartbeat data may further include other steps or processes, and the specific details of these steps or processes can be found in the prior art.

[0066] The embodiments described above are for illustrative purposes to illustrate embodiments of the present invention, but the present invention is not limited thereto, and appropriate modifications may be made based on the embodiments described above. For example, the embodiments described above may be used individually, or a combination of several of the embodiments described above may be used.

[0067] As can be seen from the above-described embodiment, bandpass filtering feature data and / or frequency spectral energy feature data are acquired based on the wireless radar signal, the time-domain bandpass filtering feature data and / or frequency-domain frequency spectral energy feature data are input into a neural network-based detection model, waveform data after detection is acquired using the detection model, and the heartbeat interval and heart rate of the target can be acquired based on the waveform data. This enables the accurate acquisition of heartbeat data non-contact and the monitoring of resting heart rate and heart rate variability, making heartbeat monitoring convenient at any time, providing a comfortable user experience, and offering high accuracy.

[0068] <Example of the second aspect> An embodiment of the present invention provides a device for extracting heartbeat data based on a wireless radar signal, and the same details as those described in the embodiment of the first aspect are omitted here.

[0069] Figure 11 shows an apparatus for extracting heartbeat data based on a wireless radar signal in an embodiment of the present invention. As shown in Figure 11, the apparatus 1100 for extracting heartbeat data based on a wireless radar signal includes the following:

[0070] Collection unit 1101; collects radio radar signals detected by the radar; Filtering unit 1102: Performs bandpass filtering of a first frequency band on the acquired radio radar signal, wherein the first frequency band is a predetermined frequency band related to the heartbeat of the target to be detected; Calculation unit 1103: Calculates the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; Feature extraction unit 1104: Performs feature extraction on the data after bandpass filtering to obtain bandpass filtering feature data, and / or performs feature extraction on the frequency spectral energy at each time point within the second frequency band to obtain frequency spectral energy feature data; Detection unit 1105: Inputs time-domain bandpass filtering feature data and / or frequency-domain frequency spectral energy feature data into a neural network-based detection model, and uses the detection model to acquire the detected waveform data; and Acquisition unit 1106: Obtains the heartbeat interval and heart rate of the target to be detected based on the waveform data.

[0071] In some embodiments, the calculation unit 1103 obtains the frequency spectral energy at each time point by converting the data after bandpass filtering from a time-domain signal to a frequency-domain signal using a sliding time window.

[0072] In some embodiments, the computing unit 1103 uses a short-time Fourier transform or wavelet transform to convert the bandpass filtered data from a time-domain signal to a frequency-domain signal.

[0073] In some embodiments, the frequency spectral energy feature data includes one or more of the following: Data obtained after summing the frequency spectral energies at each time point for a portion of the frequency band within the second frequency band; Data obtained after determining the maximum value of the frequency spectral energy at each time point in a portion of the frequency band within the aforementioned second frequency band; The data obtained after summing the frequency spectral energies of all frequency bands within the second frequency band at each time point; and This data is obtained after determining the maximum value of the frequency spectral energy at each time point across all frequency bands within the aforementioned second frequency band.

[0074] In some embodiments, the bandpass filtering feature data includes one or more of the following, namely, Data after bandpass filtering has been applied to the orthogonal signals of the aforementioned radar data; Data after bandpass filtering has been performed on the in-phase signals of the aforementioned radar data; and This data is obtained after applying bandpass filtering to the aforementioned radar data to determine the amplitude values.

[0075] In some embodiments, the collection unit 1101 performs sampling at a first sampling rate, and the apparatus further includes the following: Resampling unit 1107: Before inputting the feature data into the neural network-based detection model, the feature data is resampled at a second sampling rate, of which the second sampling rate is lower than the first sampling rate; and Normalization unit 1108: Performs normalization on the feature data after resampling.

[0076] In some embodiments, the detection unit 1105 inputs feature data within a single time period into the detection model, where the single time period is a predetermined time period that ensures the inclusion of at least two R-wave information of the heartbeat.

[0077] In some embodiments, the waveform data of the detection model output is triangular wave data.

[0078] In some embodiments, the acquisition unit 1106 is further used to obtain the heartbeat interval and heart rate of the target by adding noise to the triangular wave data of the detection model output and detecting the maximum value of the triangular wave data after noise has been added.

[0079] In some embodiments, the acquisition unit 1106 is specifically used to associate the maximum values ​​of triangular wave data with the R-wave information of the electrocardiogram signal acquired by the electrocardiogram, to acquire the heartbeat interval of the target to be detected based on the time interval of adjacent maximum values, and to obtain the heart rate of the target to be detected based on the acquired heartbeat interval of the target to be detected.

[0080] Although the various components or modules according to the present invention have been described above, the present invention is not limited to these. The device 1100 for extracting heartbeat data based on a wireless radar signal may further include other components or modules, and the specific details of these components or modules can be found in the relevant technologies.

[0081] For convenience, Figure 11 only shows the connection relationships or signal directions between each component or module; however, those skilled in the art should understand that various related technologies, such as bus connections, may be employed. The above-mentioned components or modules may be implemented by hardware such as processors and memory units, but the embodiments of the present invention are not limited thereto.

[0082] The embodiments described above are for illustrative purposes to illustrate embodiments of the present invention, but the present invention is not limited thereto, and appropriate modifications may be made based on the embodiments described above. For example, the embodiments described above may be used individually, or a combination of several of the embodiments described above may be used.

[0083] As can be seen from the above-described embodiment, bandpass filtering feature data and / or frequency spectral energy feature data are acquired based on the wireless radar signal, the time-domain bandpass filtering feature data and / or frequency-domain frequency spectral energy feature data are input into a neural network-based detection model, waveform data after detection is acquired using the detection model, and the heartbeat interval and heart rate of the target can be acquired based on the waveform data. This enables the accurate acquisition of heartbeat data non-contact and the monitoring of resting heart rate and heart rate variability, making heartbeat monitoring convenient at any time, providing a comfortable user experience, and offering high accuracy.

[0084] <Example of the third side> An embodiment of the present invention provides an electronic device, which includes a device 1100 for extracting heartbeat data based on a radio radar signal as described in the embodiment of the second aspect, and the contents thereof are hereby combined. The electronic device is, for example, a computer, server, workstation, laptop computer, smartphone, etc., but the embodiments of the present invention are not limited thereto.

[0085] Figure 12 shows an electronic device in an embodiment of the present invention. As shown in Figure 12, the electronic device 1200 may include a processor (e.g., a central processor CPU) 1210 and a memory unit 1220, the memory unit 1220 being connected to the central processor 1210. The memory unit 1220 can store various types of data, and can also store a program 1221 for information processing, and can execute the program 1221 under the control of the processor 1210.

[0086] In some embodiments, the functionality of the device 1100 for extracting heartbeat data based on a radio radar signal can be integrated into a processor 1210. The processor 1210 may be configured to perform the method for extracting heartbeat data based on a radio radar signal as described in the embodiment of the first aspect.

[0087] In some embodiments, the device 1100 for extracting heartbeat data based on a wireless radar signal may be located independently of the processor 1210. For example, the device 1100 for extracting heartbeat data based on a wireless radar signal may be configured as a chip connected to the processor 1210, and the functions of the device 1100 for extracting heartbeat data based on a wireless radar signal may be realized by the control of the processor 1210.

[0088] For example, the processor 1210 may be configured to perform the following control: collect a radio radar signal that detects a target using radar; perform bandpass filtering of the acquired radio radar data in a first frequency band, the first frequency band being a predetermined frequency band relating to the heartbeat of the target; calculate the frequency spectral energy at each time point accumulated by a time window of the data after bandpass filtering; perform feature extraction on the data after bandpass filtering to obtain bandpass filtering feature data, and / or perform feature extraction on the frequency spectral energy at each time point in the second frequency band to obtain frequency spectral energy feature data; input the time-domain bandpass filtering feature data and / or frequency-domain frequency spectral energy feature data into a neural network-based detection model, and use the detection model to obtain waveform data after detection; and obtain the heartbeat interval and heart rate of the target based on the waveform data.

[0089] Furthermore, as shown in Figure 12, the electronic device 1200 may also include an input / output (I / O) device 1230, a display 1240, and the like. Since the functions of these components are similar to those in the prior art, a detailed explanation is omitted here. Note that the electronic device 1200 does not need to include all the components shown in Figure 12, and the electronic device 1200 may also include components not shown in Figure 12; for these, please refer to related technologies.

[0090] In embodiments of the present invention, a computer-readable program is further provided, wherein when the program is executed in an electronic device, the program causes the computer to perform the method of extracting heartbeat data based on the radio radar signal described in the embodiment of the first aspect in the electronic device.

[0091] In embodiments of the present invention, a storage medium storing a computer-readable program is provided, wherein the computer-readable program causes a computer to perform a method for extracting heartbeat data based on a radio radar signal described in the embodiment of the first aspect in an electronic device.

[0092] Furthermore, the above-described apparatus and method may be implemented by software or hardware, or by a combination of hardware and software. The present invention further relates to a computer-readable program as described below, that is, the program, when executed by a logic component, causes the logic component to implement the above-described apparatus or component, or to the logic component to implement the above-described various methods or steps. The logic component may be, for example, an FPGA (Field Programmable Gate Array), a microprocessor, or a processor used in a computer. The present invention further relates to a storage medium storing the above-described program, for example, a hard disk, a magnetic disk, an optical hard disk, a DVD, a flash memory, etc.

[0093] Furthermore, one or more combinations of the functional blocks shown in the drawings and / or one or more combinations of functional blocks may be implemented as a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA) or other programmable logic component, discrete gate or transistor logic component, discrete hardware assembly or any other suitable combination for performing the functions described herein. Also, one or more combinations of the functional blocks shown in the drawings and / or one or more combinations of functional blocks may further be configured as a combination of computing devices, for example, a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors connected to a DSP by communication or any other combination of any other configuration.

[0094] Furthermore, the following additional information is disclosed regarding the above-mentioned embodiments.

[0095] (Note 1) A method for extracting heartbeat data based on radio radar signals, The radar collects radio radar signals that detect an object; The acquired radio radar data is subjected to bandpass filtering of a first frequency band, wherein the first frequency band is a predetermined frequency band related to the heartbeat of the target being detected; Calculate the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; Feature extraction is performed on the data after bandpass filtering to obtain bandpass filtering feature data, and / or feature extraction is performed on the frequency spectral energy at each time point within the second frequency band to obtain frequency spectral energy feature data; The time-domain bandpass filtering feature data and / or the frequency-domain frequency spectral energy feature data are input to a neural network-based detection model, and the detected waveform data is obtained using the detection model; and A method comprising obtaining the heartbeat interval and heart rate of the target to be detected based on the waveform data.

[0096] (Note 2) The method described in Appendix 1, Calculating the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering is possible. A method comprising converting the bandpass filtered data from a time-domain signal to a frequency-domain signal using a sliding time window to obtain the frequency spectral energy at each time point.

[0097] (Note 3) The method described in Appendix 2, A method for converting the data after bandpass filtering from a time-domain signal to a frequency-domain signal using a short-time Fourier transform (STFT) or wavelet transform.

[0098] (Note 4) A method described in any one of the items in Appendix 1 to 3, The aforementioned frequency spectral energy characteristic data is Data obtained after summing the frequency spectral energies at each time point for a portion of the frequency band within the second frequency band; Data obtained after determining the maximum value of the frequency spectral energy at each time point in a portion of the frequency band within the aforementioned second frequency band; The data obtained after summing the frequency spectral energies of all frequency bands within the second frequency band at each time point; and A method comprising one or more of the data obtained after determining the maximum value of the frequency spectral energy at each time point across all frequency bands within the second frequency band.

[0099] (Note 5) A method according to any one of the items in Appendix 1 to 4, further, A method including removing a DC frequency offset from data after bandpass filtering.

[0100] (Note 6) A method described in any one of the appendices 1 to 5, The aforementioned bandpass filtering feature data is Data after bandpass filtering has been applied to the orthogonal signals of the aforementioned radar data; Data after bandpass filtering has been performed on the in-phase signals of the aforementioned radar data; and A method comprising one or more of the following: data obtained by performing bandpass filtering on the radar data to obtain amplitude values.

[0101] (Note 7) A method described in any one of the appendices 1 to 6, A method wherein the aforementioned radio radar signal is sampled based on a first sampling rate.

[0102] (Note 8) The method described in Appendix 7, Before inputting the feature data into the neural network-based detection model, the method further: The feature data is resampled based on a second sampling rate, the second sampling rate being lower than the first sampling rate; and A method that includes normalizing feature data after resampling.

[0103] (Note 9) A method described in any one of the appendices 1 to 8, A method comprising inputting feature data within a single time period into the detection model, wherein the single time period is a predetermined time (e.g., 2s) such that it includes at least two R-wave information of a heartbeat.

[0104] (Note 10) A method described in any one of the appendices 1 to 9, The waveform data output by the detection model is triangular wave data, according to the method.

[0105] (Note 11) The method described in Appendix 10, further, Noise is added to the triangular wave data output from the aforementioned detection model; and A method comprising detecting the maximum value of triangular wave data after noise has been added, and obtaining the heartbeat interval and heart rate of the target of detection.

[0106] (Note 12) The method described in Appendix 11, Detecting the maximum value of the triangular wave data after noise addition and obtaining the heartbeat interval and heart rate of the target of detection is: The aforementioned maximum value is associated with the R-wave information of the electrocardiogram signal obtained by the electrocardiogram; Obtain the heartbeat interval of the target to be detected based on the time interval between adjacent maximum values; and A method comprising obtaining the heart rate of a target based on the acquired heartbeat interval of the target.

[0107] (Note 13) Electronic equipment including memory devices and processors, The memory device stores a computer program. The processing device is an electronic device that, by executing the computer program, causes the method for extracting heartbeat data based on a radio radar signal, as described in any one of the appendices 1 to 12.

[0108] (Note 14) A storage medium containing a computer-readable program is provided. The computer-readable program causes a computer to perform a method for extracting heartbeat data based on a radio radar signal, as described in any one of the appendices 1 to 12, within an electronic device, as a storage medium.

[0109] Although preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and any modification to the present invention falls within the technical scope of the present invention as long as it does not deviate from the spirit of the invention.

Claims

1. A device for extracting heartbeat data based on wireless radar signals, A collection unit that collects wireless radar signals detected by radar; A filtering unit that performs bandpass filtering of a first frequency band on collected radio radar signals, wherein the first frequency band is a predetermined frequency band related to the heartbeat of the object to be detected; A computing unit that calculates the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; A feature extraction unit that performs feature extraction on the data after bandpass filtering to obtain bandpass filtering feature data, and / or performs feature extraction on the frequency spectral energy at each time point within the second frequency band to obtain frequency spectral energy feature data; A detection unit that inputs the time-domain bandpass filtering feature data and / or the frequency-domain frequency spectral energy feature data into a neural network-based detection model and acquires the detected waveform data using the detection model; and Includes an acquisition unit that obtains the heartbeat interval and heart rate of the target to be detected based on the waveform data, The waveform data output from the aforementioned detection model is triangular wave data. The acquisition unit is a device that adds noise to the triangular wave data output from the detection model, detects the maximum value of the triangular wave data after noise addition, and acquires the heartbeat interval and heart rate of the target of detection.

2. The apparatus according to claim 1, The calculation unit is a device that converts the data after bandpass filtering from a time-domain signal to a frequency-domain signal using a sliding time window, and obtains the frequency spectral energy at each time point.

3. The apparatus according to claim 2, The aforementioned computing unit is a device that converts the data after bandpass filtering from a time-domain signal to a frequency-domain signal using a short-time Fourier transform or a wavelet transform.

4. The apparatus according to claim 1, The aforementioned frequency spectral energy characteristic data is Data obtained after calculating the sum of the frequency spectral energies at each time point in a portion of the frequency band within the aforementioned second frequency band; Data obtained after determining the maximum value of the frequency spectral energy at each time point in a portion of the frequency band within the aforementioned second frequency band; The data obtained after calculating the sum of the frequency spectral energies at each time point across all frequency bands within the second frequency band; and An apparatus comprising one or more of the data obtained after determining the maximum value of the frequency spectral energy at each time point across all frequency bands within the second frequency band.

5. The apparatus according to claim 1, The aforementioned bandpass filtering feature data is Data after bandpass filtering has been applied to the orthogonal signals of radar data; Data after bandpass filtering has been performed on the in-phase signals of the aforementioned radar data; and A device comprising one or more of the data obtained by performing bandpass filtering on the aforementioned radar data to determine the amplitude.

6. The apparatus according to claim 1, The collection unit performs sampling at the first sampling rate, and the device, A resampling unit that resamples feature data at a second sampling rate before inputting the feature data to the neural network-based detection model, wherein the second sampling rate is lower than the first sampling rate; and The apparatus further includes a normalization unit that performs normalization on the feature data after resampling.

7. The apparatus according to claim 1, The detection unit inputs feature data within a single time period into the detection model, wherein the single time period is a predetermined time period that can be ensured to include at least two R-wave information of a heartbeat.

8. The apparatus according to claim 1, The acquisition unit is a device that associates the maximum value with the R-wave information of the electrocardiogram signal acquired by the electrocardiogram, acquires the heartbeat interval of the target to be detected based on the time interval of adjacent maximum values, and obtains the heart rate of the target to be detected based on the acquired heartbeat interval of the target to be detected.

9. A method for extracting heartbeat data based on a radio radar signal, performed by a computer, The radar collects radio radar signals that detect an object; The collected radio radar data is subjected to bandpass filtering of a first frequency band, the first frequency band being a predetermined frequency band related to the heartbeat of the detected object; Calculate the frequency spectral energy at each time point, accumulated by the time window, of the data after bandpass filtering; Feature extraction is performed on the data after the bandpass filtering to obtain bandpass filtering feature data, and / or feature extraction is performed on the frequency spectral energy at each time point within the second frequency band to obtain frequency spectral energy feature data; The bandpass filtering feature data in the time domain and / or the frequency spectral energy feature data in the frequency domain are input to a neural network-based detection model, and the detected waveform data is obtained using the detection model; and This includes obtaining the heartbeat interval and heart rate of the target to be detected based on the waveform data, The waveform data output from the aforementioned detection model is triangular wave data. Obtaining the heartbeat interval and heart rate of the target to be detected based on the waveform data is a method that involves adding noise to the triangular wave data output from the detection model, detecting the maximum value of the triangular wave data after noise addition, and obtaining the heartbeat interval and heart rate of the target to be detected.

Images