A non-contact heart rate variability monitoring method based on frequency modulated continuous wave radar

Abstract

The present invention discloses a non-contact heart rate variability monitoring method based on a frequency-modulated continuous wave radar. The specific method includes transmitting a periodic linearly frequency-modulated wave to the area where the monitoring object is located and collecting the radar echo signal; preprocessing the radar echo signal by using mixing and FFT in sequence; then extracting the position of the heart part of the monitoring object; adopting a phase correlation method to extract the motion information of the heart part of the monitoring object and calculating its acceleration signal accordingly; smoothing the obtained acceleration signal and estimating the position of the segmentation point between each heartbeat by using a peak detection method; generating an acceleration template signal for a single heartbeat and precisely segmenting the acceleration signal to obtain the inter-beat interval of each heartbeat, and calculating the heart rate variability index. The present invention uses microwaves as the detection medium and has many advantages such as strong anti-interference ability, strong penetration ability for non-metallic obstacles, high integration, low power consumption, high measurement accuracy, and strong robustness.

Description

Technical Field

[0001] The present invention relates to the field of transforming traditional industries with high - tech and non - contact vital sign monitoring, and particularly to the design of a specific non - contact heart rate variability monitoring algorithm based on frequency - modulated continuous - wave radar. Background Art

[0002] In recent years, the research on non - contact physiological feature monitoring has been in the ascendant, and various non - contact physiological feature monitoring methods have emerged continuously. Their monitoring media are diverse, including infrared, acoustic waves, and optics, etc. However, the non - contact physiological feature monitoring technologies based on the above - mentioned media all have some obvious disadvantages. For example, the monitoring method based on infrared sensors is affected by heat sources near the monitored object; the monitoring method based on ultrasonic sensors is affected by air humidity and materials such as cotton yarn that can absorb acoustic waves; the optical - based sensors are affected by factors such as lighting conditions, smoke, and body occlusion of the monitored object. The above - mentioned disadvantages limit the popularization and application of these non - contact physiological feature monitoring methods.

[0003] On the other hand, heart rate variability (HRV) is a very important vital sign index of the human body, which contains information on the regulation of the cardiovascular system by neuro - humoral factors and has a very close relationship with physiological signs such as emotional arousal, sympathetic nerve tension, and balance. In the prior art, there are already various monitoring means to monitor the heartbeat signal of the human body and thus analyze the heart rate variability index. However, the analysis of the heart rate variability index requires high measurement accuracy; and compared with physiological features such as respiration, the intensity of the heartbeat signal is very weak and is extremely vulnerable to interference. Therefore, most of the existing heart rate variability monitoring means are still based on contact - type electrocardiogram monitors or finger - clip pulse wave sensors, and the wearing of finger clips, electrodes, etc. during the measurement process will significantly reduce the user experience. Summary of the Invention

[0004] In order to solve the problems of low measurement accuracy, limited use conditions, and susceptibility to environmental factor interference in the existing non - contact physiological feature monitoring methods, and at the same time optimize the user experience of the traditional heart rate variability monitoring method in a non - contact manner, the present invention proposes a non - contact heart rate variability monitoring method based on frequency - modulated continuous - wave radar. Through the innovative design of the radar echo signal processing algorithm, the signal - to - noise ratio of the heartbeat - related signal in the echo signal is improved, and finally a better heart rate variability monitoring effect is achieved.

[0005] The specific technical solutions adopted by the present invention are as follows:

[0006] The present invention provides a non - contact heart rate variability monitoring method based on frequency - modulated continuous - wave radar, specifically as follows:

[0007] S1: Transmit a periodic linear frequency modulated wave to the area where the monitoring object is located, and use the receiving antenna to collect the corresponding radar echo signal;

[0008] S2: Preprocess the collected radar echo signal by using mixing and fast Fourier transform in sequence;

[0009] S3: Extract the position of the heart part of the monitoring object from the preprocessed radar echo signal;

[0010] S4: Adopt the method of phase correlation to extract the motion information of the heart part of the monitoring object from the position obtained in step S3, and calculate its acceleration signal accordingly;

[0011] S5: Smooth the obtained acceleration signal by calculating the short-time average power, and estimate the position of the segmentation point between each heartbeat by the method of peak detection;

[0012] S6: Generate an acceleration template signal for a single heartbeat based on the position of the segmentation point between each heartbeat estimated in step S5; Subsequently, use the acceleration template signal to precisely segment the acceleration signal obtained in step S4 to obtain the inter-beat intervals of each heartbeat;

[0013] S7: Calculate the heart rate variability index based on the inter-beat intervals obtained in step S6, and further realize non-contact heart rate variability monitoring.

[0014] Preferably, in step S1, the pulse frequency modulation range of the linear frequency modulated wave is 77 GHz to 81 GHz, the duration of a single pulse is 50 us; the sampling frequency of the fast time axis is 4 MHz, and the number of sampling points is 128; the sampling frequency of the slow time axis is 100 Hz.

[0015] Preferably, in step S2, the output after being processed by the mixing method is an intermediate frequency signal; the amplitude of this intermediate frequency signal is a constant, and the frequency and phase are both proportional to the distance of the reflecting object relative to the receiving antenna array.

[0016] Preferably, step S3 is specifically as follows:

[0017] S3-1: For the nth radar echo signal received, extract the phase of its kth element after being preprocessed in step S2 ;

[0018] S3-2: Select a time window with a length of T, and process it in sequence according to the method in step S3-1 to obtain the sequence Φ k :

[0019]

[0020] S3-3: Unwind the obtained sequence Φ k and calculate its spectrum through fast Fourier transform, and statistically calculate its energy E in the heartbeat frequency band [0.8Hz, 2Hz] k ;

[0021] S3-4: Traverse k to find the maximum value of E k and determine the distance of the heart part of the monitored object relative to the radar antenna accordingly, which is the position where the heart part of the monitored object is located.

[0022] Preferably, the step S4 is specifically as follows:

[0023] Extract the phase sequence reflecting the distance change between the heart part of the monitored object and the receiving antenna array according to the formula After unwinding, obtain the corresponding body movement signal sequence {R }; where, λ is the wavelength of the transmitted frequency-modulated continuous pulse; m

[0024] Subsequently, through the formula

[0025]

[0026] obtain the corresponding acceleration signal sequence {a m}; where, Δt is the sampling interval of the slow time axis, R m+1 is the (m + 1)-th term in the body movement signal sequence {R m}, R m is the m-th term in the body movement signal sequence {R m}, and R m-1 is the (m - 1)-th term of the body movement signal sequence {R m}.

[0027] Preferably, the step S5 is specifically as follows:

[0028] S5-1: Define the average power P L (n0) of the acceleration signal within the neighborhood of length L at time n0 as

[0029]

[0030] where, L is half of the time window length selected for calculating the average power, and a i is the i-th term in the acceleration signal sequence {a m};

[0031] Take L1 as the number of sampling points corresponding to half of the single heartbeat cycle duration, and take L2 = 0.4L1; Process the acceleration signal sequence {a m} term by term according to the following formula to obtain the smoothed heartbeat signal sequence {Y m};​

[0032]

[0033] Among them, L1 and L2 represent different time window lengths selected when calculating the average power. represents the average power of the acceleration signal within the neighborhood of length L1 at time m; represents the average power of the acceleration signal within the neighborhood of length L2 at time m;

[0034] S5-2: For the smoothed heartbeat signal sequence {Y m}, find the subscripts t corresponding to all its maximum points in the sequence through peak detection

[0035] t = [t1, t2,... t i ..., t I ,

[0036] The obtained t i is the subscript corresponding to the segmentation point between each heartbeat in the sequence; where I represents the total number of maximum points found.

[0037] Preferably, the step S6 is specifically as follows:

[0038] S6-1: Based on the estimated value t of the position of the segmentation point of each heartbeat cycle obtained in step S5, segment the acceleration signal sequence {a m} to obtain the following (I - 1) data segments:

[0039]

[0040] S6-2: Average the data segments obtained in step S6-1 to obtain the heartbeat acceleration template signal Template:

[0041]

[0042] During the calculation, interpolation is used to ensure that the segments of each acceleration signal in S6-1 have the same length;

[0043] S6-3: Select a new set of segmentation points S:

[0044] S = [s1, s2, ……, s I ,

[0045] to obtain a new truncation of the acceleration signal sequence {a m};

[0046]

[0047] S6-4: Define the matching degree of the new truncated signal and the acceleration template signal obtained in step S6-3 through the following formula:

[0048]

[0049] During the calculation, interpolation is used to ensure that each is the same length as the Template;

[0050] S6-5: Repeat steps S6-3 to S6-4 to calculate the matching degree of the truncated signal and the acceleration template signal corresponding to each set of segmentation point sets, and find the set of segmentation point sets S with the smallest D s as the optimal segmentation under the current template signal; Based on S s Generate a new template signal according to step S6-2;

[0051] S6-6: Repeat the process described in S6-5 until D calculated through step S6-4 in a certain repetition process is less than a pre-set threshold. At this time, the segmentation point set S is the position of the segmentation point between each heartbeat finally obtained.

[0052] The present invention has the following beneficial effects compared with the prior art:

[0053] The present invention uses microwave as the detection medium, which can effectively avoid the problems of complex usage methods, poor usage experience and limited user groups of traditional contact means, and has many advantages such as strong anti-interference ability, strong penetration ability for non-metallic obstacles, high integration, low power consumption, high measurement accuracy and strong robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is the technical flow chart of the monitoring method of the present invention;

[0055] Figure 2 is the specific step flow chart of the monitoring method of the present invention;

[0056] Figure 3 is the display diagram of the measurement effect of each inter-beat interval in the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0057] The present invention will be further described and explained below in conjunction with the drawings and specific embodiments. The technical features of each embodiment in the present invention can be combined correspondingly without conflict.

[0058] As Figure 1 shown, the present invention provides a non-contact heart rate variability monitoring method based on a frequency-modulated continuous-wave radar. The method mainly includes the following processes: preprocessing of the radar echo signal, determination of the heart position of the monitoring object, extraction of the heartbeat signal, segmentation between each heartbeat, and calculation of heart rate variability indexes. The following will be based onFigure 2 Describe the specific steps of this method.

[0059] S1: Use the transmitting antenna to transmit a periodic linear frequency modulated wave to the area where the monitoring object is located, and use the receiving antenna to collect the corresponding radar echo signal.

[0060] In practical applications, the pulse frequency modulation range of the linear frequency modulated wave is 77 GHz to 81 GHz, and the duration of a single pulse is 50 μs; the sampling frequency of the fast time axis is 4 MHz, and the number of sampling points is 128; the sampling frequency of the slow time axis is 100 Hz.

[0061] S2: Preprocess the collected radar echo signal by using mixing and fast Fourier transform (FFT) in sequence. Among them, the output after the mixing process is an intermediate frequency signal; the amplitude of this intermediate frequency signal is a constant, and both the frequency and the phase are proportional to the distance of the reflecting object relative to the receiving antenna array. The specific steps are as follows:

[0062] The mixing output is an intermediate frequency signal x out ,

[0063]

[0064] The amplitude A of this intermediate frequency signal is a constant, and both the frequency and the phase are proportional to the distance R between the reflecting surface and the antenna, as shown in the following formula:

[0065]

[0066] Where S is the rate of change of the linear frequency modulated continuous wave frequency with time, c is the speed of light, and λ is the wavelength of the used frequency modulated continuous wave, all of which are constants. Finally, perform FFT on the intermediate frequency signal output by mixing to obtain the components of different frequencies in the spectrum, which represent the radar echo signals reflected by objects at different distances.

[0067] S3: Extract the position of the heart part of the monitoring object from the preprocessed radar echo signal. The specific steps are as follows:

[0068] S3-1: For the nth radar echo signal received, extract the phase of its kth element after preprocessing in step S2

[0069] S3-2: Select a time window with a length of T, and process it in sequence according to the method in step S3-1 to obtain the sequence Φ k :

[0070]

[0071] S3-3: Perform the following operations on the obtained sequence Φ kUnwind it and calculate its spectrum through the Fast Fourier Transform, and statistically calculate the energy E in the heartbeat frequency band [0.8Hz, 2Hz]. k ;

[0072] S3-4: Traverse k to find the maximum value of E k and determine the distance of the heart part of the monitored object relative to the radar antenna accordingly, which is the position where the heart part of the monitored object is located.

[0073] S4: Adopt the method of phase correlation to extract the motion information of the heart part of the monitored object from the position obtained in step S3, and calculate its acceleration signal accordingly. The specific steps are as follows:

[0074] According to the formula extract the phase sequence reflecting the change in the distance between the heart part of the monitored object and the receiving antenna array After unwinding, obtain the corresponding body motion signal sequence {R m}; where λ is the wavelength of the transmitted frequency-modulated continuous pulse.

[0075] Subsequently, through the formula

[0076]

[0077] obtain the corresponding acceleration signal sequence {a m}; where Δt is the sampling interval of the slow time axis.

[0078] The above steps S3 and S4 are unified as extracting the heartbeat information of the monitored object. In practical applications, for example, the spectrum obtained after the nth radar echo signal received by the receiving antenna undergoes the above preprocessing is a complex sequence, denoted as this sequence as x n , and the kth element in this sequence is denoted as x n,k . The phase of this complex element can be obtained through the arctangent function

[0079]

[0080] where Im(x n,k ) represents the imaginary part of the complex number x n,k , and Re(x n,k ) represents the real part of the complex number x n,k .

[0081] Select the time window length as T, and calculate the spectrum of each radar echo signal within the time window length T according to the above preprocessing method

[0082] x n+1 ,x n+2 ,……,x n+T

[0083] Select the k-th element from each spectrum sequence respectively, and calculate its phase according to the aforementioned method Then form a new sequence Φ k

[0084]

[0085] Perform FFT on this sequence to obtain its spectrum, and calculate its total energy E in the heartbeat frequency band [0.8Hz, 2Hz] accordingly k ; By traversing the values of k in the above process, the energy values E of objects at different distances within the monitoring range in the heartbeat frequency band can be obtained

[0086] E = [E1, E2, ……, E N

[0087] where N represents the number of points of the spectrum obtained after preprocessing. Find the largest item in E (for example, the p-th item), which represents the maximum energy in the heartbeat frequency band at this distance. It is considered that the subscript p corresponding to this maximum item E p is the position of the heart of the monitored object. Select the p-th item from each preprocessed spectrum and extract its phase according to the above arctangent function to obtain a phase sequence

[0088]

[0089] After unwrapping the above phase sequence, the body movement signal sequence {R m}, m = 1, 2, ……, T;

[0090] The extraction of the heartbeat information of the monitored object is realized by calculating the second-order difference signal of the body movement signal sequence {R m}. For the body movement signal sequence {R m} obtained above, calculate its acceleration signal sequence {a m} item by item according to the following formula

[0091]

[0092] where Δt is the sampling interval of the slow time axis

[0093] S5: Although the amplitude of the respiratory signal in the body movement signal is larger than that of the heartbeat signal, its change process is relatively slow. Therefore, by calculating the acceleration signal of the body movement signal, the respiratory signal is suppressed and the heartbeat signal is enhanced. Calculate the short-time average power to smooth the obtained acceleration signal, and estimate the position of the segmentation point between each heartbeat by the method of peak detection

[0094] This step is specifically as follows ​

[0095] Segmentation of the heartbeat signal is first performed by calculating the power to smooth the obtained acceleration signal sequence {a m}. The average power at time n0 is defined as

[0096]

[0097] where L is half of the time window length for calculating the average power.

[0098] Process the acceleration signal sequence {a m} item by item according to the following formula to obtain the smoothed heartbeat signal sequence {Y m}:

[0099]

[0100] where L1 and L2 represent different time window lengths selected when calculating the average power. Take L1 as the number of sampling points corresponding to half of the duration of a single heartbeat cycle (e.g., 0.8 - 1.2 s), and take L2 = 0.4L1;

[0101] The smoothed heartbeat signal sequence {Y m} is a quasi - sine signal. By finding the peaks in the sequence {Y m} and recording the subscripts of each peak point in the sequence:

[0102] t = [t1, t2, ……, t I

[0103] where I represents the number of peak points in the sequence {Y m}. t i is the subscript of the segmentation point between each estimated heartbeat in the sequence. Based on this, the inter - heartbeat intervals of each heartbeat can be obtained.

[0104] S6: However, when performing the above - mentioned smoothing process, the calculation of the average power over a period of time will cause the loss of some details in the acceleration signal sequence {a m}, resulting in a large error in the inter - heartbeat intervals obtained based on the above - mentioned heartbeat segmentation points and not meeting the monitoring requirements of the heart rate variability index. Therefore, it is necessary to perform further refined segmentation processing on the acceleration signal sequence {a m}. That is, based on the positions of the segmentation points between each estimated heartbeat in step S5, generate the acceleration template signal of a single heartbeat; then use this acceleration template signal to precisely segment the acceleration signal obtained in step S4 to obtain the inter - heartbeat intervals of each heartbeat.

[0105] The specific steps are as follows:

[0106] ​Based on the positions \(t\) of the heartbeat segmentation points obtained from the above estimations, truncate the acceleration signal sequence \(\{a\ m \}\) into the following segments

[0107]

[0108] where represents the sequence of the acceleration signal sequence \(\{a\ m \}\) corresponding to the two segmentation points \([t_1, t_2]\).

[0109] Average the above acceleration signal segments to obtain the template signal Template of the heartbeat acceleration signal

[0110]

[0111] For acceleration signal segments with different lengths, it is necessary to unify their lengths by interpolation.

[0112] Use the template signal Template to re-segment the original acceleration signal sequence \(\{a\ m \}\). Select a set of segmentation points \(S = [s_1, s_2, \cdots, s m \) in the original acceleration signal sequence \(\{a\ I \}\). Obtain a new truncation of the acceleration signal sequence \(\{a\ m \}\)

[0113]

[0114] Define the matching degree between the new truncation and the template by the following formula

[0115]

[0116] When a certain truncation has a different length from the template Template, still make the length of this truncation consistent with the length of the template by interpolation.

[0117] Repeatedly adjust the segmentation points \(S\) until the value of \(D\) reaches the minimum under the current template signal; denote the segmentation point set at this time as \(S_1\), generate a new template signal based on \(S_1\) according to the above method, denoted as Template1, and re-segment the original acceleration signal sequence \(\{a\ m \}\) based on Template1. Until the calculated \(D\) is less than the pre-set threshold, it is considered that the segmentation of the current acceleration signal sequence \(\{a\ m \}\) is completed. The segmentation point set at this time

[0118] \(S=\{s_1, s_2, \cdots s I \}\)

[0119] That is the position of the segmentation point between each heartbeat finally.

[0120] S7: Based on the inter-beat intervals obtained in step S6, calculate the heart rate variability index, thereby realizing non-contact heart rate variability monitoring.

[0121] Specifically, calculate the inter-beat interval RR for each heartbeat accordingly

[0122] RR = [s2 - s1, s3 - s2, ……, s I -s I-1

[0123] Based on the above inter-beat interval RR, calculate heart rate variability related indexes such as MEAN, SDNN, etc. according to the following formula

[0124]

[0125]

[0126] Embodiment

[0127] To make the technical means, creative features, achieved purposes and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0128] A non-contact heart rate variability monitoring method based on a frequency-modulated continuous wave radar is proposed in this embodiment, specifically as follows:

[0129] S1: Transmit a periodic linearly frequency-modulated wave to the area where the monitoring object is located, and use an antenna to collect the corresponding echo signal. The linear frequency modulation range of the transmitted linearly frequency-modulated wave is 77 - 81 GHz, and the single pulse duration period is 50 us; the fast time axis sampling frequency is 4 MHz, and the number of sampling points is 128; the slow time axis sampling frequency is 100 Hz.

[0130] S2: Preprocess the radar echo signal by using mixing and FFT. The mixing output is an intermediate frequency signal x out .

[0131]

[0132] The amplitude A of this intermediate frequency signal is a constant, and the frequency and phase are both proportional to the distance R between the reflecting surface and the antenna, as shown in the following formula:

[0133]

[0134] Where S is the rate of change of the linearly frequency-modulated continuous wave frequency with time, c is the speed of light, and λ is the wavelength of the used frequency-modulated continuous wave. In this example, S = 70 MHz / μs and λ = 4 mm.

[0135] ​Finally, perform FFT on the intermediate-frequency signal output from the mixer. The components of different frequencies in the spectrum represent the radar echo signals reflected by objects at different distances.

[0136] S3 Extract the position of the heart part of the monitoring object from the radar echo signal, specifically the distance relative to the antenna.

[0137] S3-1 For the nth radar echo signal received, after preprocessing, extract its kth element x through the following arctangent function n,k phase

[0138]

[0139] where Im(x n,k ) represents the imaginary part of the complex number x n,k , and Re(x n,k ) represents the real part of the complex number x n,k .

[0140] S3-2 Select a time window of length T and sequentially process all the radar echo signals in the time window according to the method in S3-1 to obtain a new sequence Φ k

[0141]

[0142] S3-3 Unwrap the sequence Φ in S3-2 and calculate its spectrum through FFT, and statistically calculate its energy E in the heartbeat frequency band [0.8Hz, 2Hz] k k

[0143] S3-4 Traverse k and find the maximum value of E k , and accordingly determine the distance of the heart part of the monitoring object relative to the radar antenna.

[0144] S4 According to Extract the phase sequence reflecting the change in the distance between the heart part of the monitoring object and the antenna array After unwrapping, obtain the corresponding motion information sequence {R m}; and through

[0145]

[0146] Obtain the corresponding acceleration signal sequence {a m}

[0147] where Δt is the sampling interval of the slow time axis

[0148] S5 smooths the acceleration signal in S4 by calculating the short-time average power, and estimates the positions of the segmentation points between each heartbeat through peak detection

[0149] S5-1 takes L1 = 50, L2 = 20, and through

[0150]

[0151]

[0152] smoothes the acceleration signal sequence {a m} obtained in S4 to obtain the smoothed heartbeat signal sequence {Y m}, and {Y m} should be a quasi-sinusoidal signal

[0153] S5-2 For the smoothed heartbeat signal sequence {Y m}, find all the subscripts t corresponding to its maximum points in the sequence through peak detection

[0154] t = [t1, t2, ……, t I

[0155] It is considered that t i is the subscript corresponding to the segmentation point between each heartbeat, where I represents the number of peak points in the sequence {Y m}

[0156] S6 generates an acceleration template signal for a single heartbeat based on the segmentation points between each heartbeat estimated in S5, and uses this template signal to precisely segment the acceleration signal in S4 to obtain the inter-beat intervals of each heartbeat

[0157] S6-1 Based on the estimated values t of the segmentation points of each heartbeat cycle obtained in S5, segment the acceleration signal sequence {a m} to obtain the following I - 1 data segments

[0158]

[0159] S6-2 Averages the acceleration signal segments in S6-1 to obtain the template signal Template of the heartbeat acceleration signal

[0160]

[0161] Ensures that the segment lengths of each acceleration signal sequence {a m} in S6-1 are consistent through interpolation

[0162] S6-3 Selects a new set of segmentation points S

[0163] ​S = [s1, s2, ……, s I

[0164] Obtain a new truncation of the acceleration signal sequence {a m}}

[0165]

[0166] S6-4 defines the matching degree between the new truncation and the template through the following formula

[0167]

[0168] Ensure that each truncation has the same length as the template Template through interpolation

[0169] S6-5:

[0170] Repeat the process described in S6-3 to S6-4 to calculate the matching degree between the truncation corresponding to each set of segmentation point sets and the acceleration template signal, and find the set of segmentation point sets S with the smallest D s as the optimal segmentation under the current template signal. Based on S s Generate a new template signal according to the process described in S6-2

[0171] S6-6: Repeat the process described in S6-5 until D calculated in S6-4 in a certain repetition process is less than a preset threshold. At this time, the segmentation point set S is the position of the segmentation point between each heartbeat finally obtained. Calculate the inter-beat interval of each heartbeat based on this segmentation point set S

[0172] S7 calculates the heart rate variability index according to the heart rate variability definition formula based on the inter-beat intervals obtained in S6. Taking the MEAN index as an example

[0173]

[0174] Adopt the method of the above embodiment to monitor the heart rate variability of the monitored object, and obtain the following results

[0175] Figure 3 For the display of the measurement effect of each inter-beat interval, the red curve in the figure represents the inter-beat intervals obtained from the electrocardiogram signal collected by a computerized electrocardiograph, which is the reference signal; the blue curve represents the inter-beat intervals RR measured by a frequency modulated continuous wave radar based on the method proposed in this patent

[0176] RR = [s2 - s1, s3 - s2, ……, s I - s I-1

[0177] ​​​Table 1 shows the measurement results of the MEAN index of heart rate variability in 6 groups of experimental data with a duration of 5 minutes.

[0178] Table 1

[0179]

[0180] The embodiments described above are only a preferred solution of the present invention, but they are not intended to limit the present invention. Those of ordinary skill in the relevant technical fields can still make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all technical solutions obtained by means of equivalent replacement or equivalent transformation fall within the protection scope of the present invention.

Claims

1. A non-contact heart rate variability monitoring method based on a frequency modulated continuous wave radar, characterized in that, The details are as follows: S1: Transmit a periodic linear frequency modulated wave to the area where the monitoring object is located, and use the receiving antenna to collect the corresponding radar echo signal; S2: Preprocess the collected radar echo signal by means of mixing and fast Fourier transform in sequence; S3: Extract the position of the heart part of the monitoring object from the preprocessed radar echo signal; S4: Adopt the method of phase correlation to extract the motion information of the heart part of the monitoring object from the position obtained in step S3, and calculate its acceleration signal accordingly; S5: Smooth the obtained acceleration signal by calculating the short-time average power, and estimate the position of the segmentation point between each heartbeat by the method of peak detection; S6: Based on the position of the segmentation point between each heartbeat estimated in step S5, generate an acceleration template signal for a single heartbeat; Subsequently, use the acceleration template signal to precisely segment the acceleration signal obtained in step S4 to obtain the inter-beat interval of each heartbeat; S7: Based on the inter-beat interval obtained in step S6, calculate the heart rate variability index, and further realize non-contact heart rate variability monitoring; The specific content of step S3 is as follows: S3-1: For the nth radar echo signal received, after preprocessing in step S2, extract the phase of its kth element S3-2: Select a time window of length T and process it successively according to the method in step S3-1 to obtain the sequence Φ k : S3-3: Unwind the obtained sequence Φ k and calculate its spectrum through fast Fourier transform, and then count its energy E in the heartbeat frequency band [0.8Hz, 2Hz] k ; S3-4: Traverse k to find the maximum value of E k and determine the distance of the heart part of the monitoring object relative to the radar antenna accordingly, which is the position where the heart part of the monitoring object is located; The specific content of step S6 is as follows: S6-1: Based on the estimated value t of the position of each heart beat cycle segmentation point obtained in step S5, the acceleration signal sequence {a m} is segmented to obtain the following (I-1) data segments: S6-2: Average the data segment obtained in step S6-1 to obtain the heartbeat acceleration template signal Template: Interpolate during calculation to ensure that the segments of each acceleration signal in S6-1 have the same length; S6-3: Select a new set of segmentation points S: S = [s1, s2, ……, s I , Obtain a new truncation of the acceleration signal sequence {a m}: S6-4: Define the matching degree between the new truncation obtained in step S6-3 and the acceleration template signal through the following formula: During calculation, interpolation is used to ensure that each is the same length as the Template; S6-5: Repeat steps S6-3 to S6-4 to calculate the matching degree between the truncation corresponding to each set of segmentation points and the acceleration template signal, and find the set of segmentation points S with the smallest D s as the optimal segmentation under the current template signal; Based on S s Generate a new template signal according to step S6-2; S6-6: Repeat the process of S6-5 until D calculated through step S6-4 in a certain repetition process is less than a preset threshold. At this time, the set of segmentation points S is the position of the segmentation point between each heartbeat finally obtained.

2. The non-contact heart rate variability monitoring method based on a frequency modulated continuous wave radar according to claim 1, wherein In step S1, the pulse frequency modulation range of the linear frequency modulated wave is 77 GHz to 81 GHz, and the duration of a single pulse is 50 us; the sampling frequency of the fast time axis is 4 MHz, and the number of sampling points is 128; the sampling frequency of the slow time axis is 100 Hz.

3. The non-contact heart rate variability monitoring method based on a frequency modulated continuous wave radar according to claim 1, characterized in that, In step S2, the output after being processed by the mixing method is an intermediate frequency signal; the amplitude of this intermediate frequency signal is a constant, and the frequency and phase are both proportional to the distance of the reflecting object relative to the receiving antenna array.

4. The non-contact heart rate variability monitoring method based on a frequency modulated continuous wave radar according to claim 1, characterized in that, The specific content of step S4 is as follows: According to the formula Extract the phase sequence that reflects the change in the distance between the heart part of the monitoring object and the receiving antenna array After unwrapping, obtain the corresponding body movement signal sequence {R m}; where λ is the wavelength of the transmitted frequency-modulated continuous pulse Subsequently, through the formula Obtain the corresponding acceleration signal sequence {a m}; where, Δt is the sampling interval of the slow time axis, and R m+1 is the (m + 1)-th term of the body movement signal sequence {R m}, R m is the m-th term of the body movement signal sequence {R m}, and R m-1 is the (m - 1)-th term of the body movement signal sequence {R m}.

5. The non-contact heart rate variability monitoring method based on a frequency modulated continuous wave radar according to claim 1, wherein The specific content of step S5 is as follows: S5-1: Define the average power P of the acceleration signal within a neighborhood of length L at time n0 L (n0) as where L is half of the time window length for calculating the average power, and a i is the i-th term in the acceleration signal sequence {am}; Take L1 as the number of sampling points corresponding to half of the single heartbeat cycle duration, and take L2 = 0.4L1; according to the following formula, process the acceleration signal sequence {a m} item by item to obtain the smoothed heartbeat signal sequence {Y m}; Among them, L1 and L2 represent different time window lengths selected when calculating the average power. represents the average power of the acceleration signal in the neighborhood of the L1 length at time m; represents the average power of the acceleration signal in the neighborhood of the L2 length at time m; S5-2: For the smoothed heartbeat signal sequence {Y m}, find the subscripts t corresponding to all its maximum points in the sequence by peak detection t = [t1, t2, … t i …, t I , The obtained t i is the subscript corresponding to the segmentation point between each heartbeat in the sequence; where I represents the total number of maximum value points found.

Images