阅读说明：本技术 睡眠过程中健康指标的监测方法 (Method for monitoring health index in sleeping process ) 是由 李永川 银琪 彭飞 于 2019-09-20 设计创作，主要内容包括：本发明公开一种睡眠过程中健康指标的监测方法,采用心率波形间隔统计算法计算心率,心率波形间隔统计算法包括以下：步骤1、采用预定采样频率对原始BCG波形进行采样,获得原始BCG采样波形；步骤2、对原始BCG采样波形进行第一低通滤波,得到波形A,将原始BCG采样波形进行平滑滤波,得到波形B；步骤2、用波形A减去波形B,得到心跳信号波形C；步骤3、查找波形C中的所有波峰和波谷；步骤4、统计波形C中在预设时间内的所有波峰与波峰之间的采样点数,统计预设时间内的所有波谷与波谷之间的采样点数；等步骤。采用统计波峰波谷的距离,计算出心跳间隔,其中的无规律体动信号会被直接忽略,而有规律的峰值,会被查找到,从而计算出心率。(The invention discloses a method for monitoring health indexes in a sleeping process, which adopts a heart rate waveform interval statistical algorithm to calculate a heart rate, wherein the heart rate waveform interval statistical algorithm comprises the following steps: step 1, sampling an original BCG waveform by adopting a preset sampling frequency to obtain an original BCG sampling waveform; step 2, carrying out first low-pass filtering on the original BCG sampling waveform to obtain a waveform A, and carrying out smooth filtering on the original BCG sampling waveform to obtain a waveform B; step 2, subtracting the waveform B from the waveform A to obtain a heartbeat signal waveform C; step 3, searching all wave crests and wave troughs in the waveform C; step 4, counting the number of sampling points between all wave crests and wave crests in the waveform C within a preset time, and counting the number of sampling points between all wave troughs and wave troughs in the preset time; and the like. By counting the distance between the wave crest and the wave trough, the heartbeat interval is calculated, wherein irregular physical movement signals can be directly ignored, and regular peak values can be found out, so that the heart rate is calculated.)

1. The method for monitoring the health index in the sleeping process is characterized in that: calculating the heart rate by adopting a heart rate waveform interval statistical algorithm, wherein the heart rate waveform interval statistical algorithm comprises the following steps:

step 1, sampling an original BCG waveform by adopting a preset sampling frequency to obtain an original BCG sampling waveform;

step 2, carrying out first low-pass filtering on the original BCG sampling waveform to obtain a waveform A, and carrying out smooth filtering on the original BCG sampling waveform to obtain a waveform B;

step 2, subtracting the waveform B from the waveform A to obtain a heartbeat signal waveform C;

step 3, searching all wave crests and wave troughs in the waveform C;

step 4, counting the number of sampling points between all wave crests and wave crests in the waveform C within a preset time, and counting the number of sampling points between all wave troughs and wave troughs in the preset time;

step 5, obtaining a maximum peak value according to the number of the counted sampling points between the wave crests, wherein the maximum peak value is a value with the largest number of sampling points between the wave crests, and calculating the heart rate according to the formula (1):

heart rate 60 sample frequency/max peak (1)

The sampling frequency in equation (1) is a predetermined sampling frequency.

2. The method for monitoring health indicators during sleep as claimed in claim 1, wherein: after the step 5, the method further comprises the following steps:

step 6, carrying out second low-pass filtering on the waveform A to obtain a first respiratory waveform, removing direct-current components from the waveform A to obtain a second respiratory waveform, normalizing the second respiratory waveform, and multiplying the second respiratory waveform by a preset constant to obtain a third respiratory waveform;

and 7, searching the wave crest of the second respiratory waveform, and calculating the variation coefficient by using the size and the interval of the wave crest.

3. The method for monitoring health indicators during sleep as claimed in claim 2, wherein: after step 7, the method further comprises the following steps:

and 8, calculating the multiples of the second respiratory waveform and the third respiratory waveform, and when the coefficient of variation is less than 0.2, adjusting the amplification factor of the circuit for acquiring the original BCG signal to enable the multiples of the second respiratory waveform and the third respiratory waveform to be 0.5-2.0.

4. The method for monitoring health indicators during sleep as claimed in claim 3, wherein: after the step 8, the method further comprises the following steps:

and 9, detecting the variance of the third respiratory waveform of the last time period by adopting a variance method, setting the third respiratory waveform as apnea when the variance is smaller than a threshold value, and sending out an awakening signal when the apnea continuously reaches a preset apnea time.

5. The method of claim 4, wherein the method comprises: after step 9, the method further comprises:

and step 10, after the awakening signal is sent out, stopping the awakening signal when the apnea is not detected.

6. The method for monitoring health indicators during sleep as claimed in claim 4 or 5, wherein: the wake-up signal is a voice signal.

7. The method for monitoring health indicators during sleep as claimed in claim 4 or 5, wherein: the preset time is 20 seconds, the last time period is 5 seconds, and the preset pause time is 20 seconds.

8. The method for monitoring health indicators during sleep as claimed in any one of claims 1-5, wherein: the predetermined sampling frequency is 50.

9. The method for monitoring health indicators during sleep as claimed in claim 2, wherein: the predetermined constant is 200.

10. The method for monitoring health indicators during sleep as claimed in claim 1, wherein: step 5 also includes judging the heartbeat quality, which is as follows:

acquiring a second large peak value and a third large peak value, wherein the second large peak value is a numerical value with a second maximum of numerical values of sampling points between the wave crests, the third large peak value is a numerical value with a third maximum of numerical values of the sampling points between the wave crests, and calculating the second large peak value-the maximum peak value and the third large peak value-the second large peak value when:

and when the maximum peak value is the second large peak value, and the maximum peak value is the third large peak value, and the second large peak value, the heartbeat quality is judged to be good.

Technical Field

The invention relates to the field of nursing, in particular to a method for monitoring health indexes in a sleeping process.

Background

The heart rate, respiration, apnea and movement during sleep are important physiological indicators of the human body. The patient with the apnea can repeatedly hold back and wake up in sleep, headache after waking up, hypomnesis, reaction retardation, reduction of working capacity and the like, the heart rate, respiration, apnea, body movement duration and body movement times in the sleep process are monitored, and the sleep quality and health condition can be analyzed from the obtained data. The patient with the apnea is monitored and awakened, and the patient can be prevented from being in the apnea state for a long time. The existing breathing machine has high price, and the patient is inconvenient because the mask needs to be worn. The existing various sleep monitoring cushions and sleep pillows have poor anti-interference capability and inaccurate detection effect, and have no intervention function after apnea is detected.

In existing sleep monitoring mats, BCG signals are typically collected for analysis by piezoelectric film sensors. The signals are mixed with respiration and heartbeat signals of a human body, and respiration rate and heart rate information are obtained through algorithm processing of the signals. In the prior art, the heart rate is generally calculated by performing machine learning or fourier transform and the like on a BCG signal after high-pass filtering, and the breathing waveform obtained by low-pass filtering the BCG signal is calculated by fourier transform, but because pressure changes, such as speaking, lifting hands, scratching itch and the like, are generated on a sensor when a human body moves, the pressure changes can be collected by a piezoelectric film sensor, and the physiological indexes such as breathing and heart rate are inaccurate to detect due to the fact that the physiological indexes are mistaken for the breathing and heart rate signals during algorithm analysis, the following defects exist:

1. the prior art has the defects of large calculated amount, inaccurate calculated result and no apnea alarm intervention function.

2. In the prior art, the BCG signals are influenced to different degrees by the body weight, sleeping posture, mattress thickness and the like of a human body.

Disclosure of Invention

The invention aims to provide a method for monitoring health indexes in a sleeping process, wherein the heart rate calculation adopts the distance of the wave crests and the wave troughs of statistics to calculate the heartbeat interval, wherein existing irregular physical movement signals can be directly ignored, and existing regular peak values can be found out, so that the heart rate is calculated.

In order to achieve the purpose, the invention is realized by adopting the following technical scheme:

the invention discloses a method for monitoring health indexes in a sleeping process, which adopts a heart rate waveform interval statistical algorithm to calculate a heart rate, wherein the heart rate waveform interval statistical algorithm comprises the following steps:

step 1, sampling an original BCG waveform by adopting a preset sampling frequency to obtain an original BCG sampling waveform;

step 2, carrying out first low-pass filtering on the original BCG sampling waveform to obtain a waveform A, and carrying out smooth filtering on the original BCG sampling waveform to obtain a waveform B;

step 2, subtracting the waveform B from the waveform A to obtain a heartbeat signal waveform C;

step 3, searching all wave crests and wave troughs in the waveform C;

step 4, counting the number of sampling points between all wave crests and wave crests in the waveform C within a preset time, and counting the number of sampling points between all wave troughs and wave troughs in the preset time;

step 5, obtaining a maximum peak value according to the number of the counted sampling points between the wave crests, wherein the maximum peak value is a value with the largest number of sampling points between the wave crests, and calculating the heart rate according to the formula (1):

heart rate 60 sample frequency/max peak (1)

The sampling frequency in equation (1) is a predetermined sampling frequency.

Further, after the step 5, the method further comprises the following steps:

step 6, carrying out second low-pass filtering on the waveform A to obtain a first respiratory waveform, removing direct-current components from the waveform A to obtain a second respiratory waveform, normalizing the second respiratory waveform, and multiplying the second respiratory waveform by a preset constant to obtain a third respiratory waveform;

and 7, searching the wave crest of the second respiratory waveform, and calculating the variation coefficient by using the size and the interval of the wave crest.

Further, after step 7, the method further comprises:

and 8, calculating the multiples of the second respiratory waveform and the third respiratory waveform, and when the coefficient of variation is less than 0.2, adjusting the amplification factor of the circuit for acquiring the original BCG signal to enable the multiples of the second respiratory waveform and the third respiratory waveform to be 0.5-2.0.

Further, after the step 8, the method further comprises the following steps:

and 9, detecting the variance of the third respiratory waveform of the last time period by adopting a variance method, setting the third respiratory waveform as apnea when the variance is smaller than a threshold value, and sending out an awakening signal when the apnea continuously reaches a preset apnea time.

Further, after step 9, the method further comprises:

and step 10, after the awakening signal is sent out, stopping the awakening signal when the apnea is not detected.

Preferably, the wake-up signal is a voice signal.

Preferably, the preset time is 20 seconds, the last time period is 5 seconds, and the predetermined pause time is 20 seconds.

Preferably, the predetermined sampling frequency is 50.

Preferably, the predetermined constant is 200.

Further, step 5 further includes determining the heartbeat quality, specifically as follows:

acquiring a second large peak value and a third large peak value, wherein the second large peak value is a numerical value with a second maximum of numerical values of sampling points between the wave crests, the third large peak value is a numerical value with a third maximum of numerical values of the sampling points between the wave crests, and calculating the second large peak value-the maximum peak value and the third large peak value-the second large peak value when:

and when the maximum peak value is the second large peak value, and the maximum peak value is the third large peak value, and the second large peak value, the heartbeat quality is judged to be good.

The invention has the following beneficial effects:

1. the invention utilizes regular heartbeat signals of human body during sleeping to carry out statistics, the heart rate calculation utilizes the statistics of the distance between the wave crest and the wave trough of 20 seconds to calculate the heartbeat interval in the statistical data, wherein the existing irregular heartbeat signals can be directly ignored, and the existing regular peak values can be found out, thereby calculating the heart rate.

2. The invention calculates the variation coefficient by utilizing the wave crest of the calculated respiration waveform, when the variation coefficient is smaller, the situation has complete and non-interference respiration waveform, and the problem of different waveform amplitudes caused by the weight, the sleeping posture, the thickness of the mattress and the like of the human body can be solved by adjusting the circuit amplification factor at the moment. After the amplitude of the waveform is adjusted, the apnea can be detected by using a variance method, and the accuracy is high.

3. The invention has small calculation amount and can be very conveniently transplanted to a singlechip or an embedded system for real-time calculation.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below.

The invention discloses a method for monitoring health indexes in a sleeping process, which adopts a heart rate waveform interval statistical algorithm to calculate a heart rate, wherein the heart rate waveform interval statistical algorithm comprises the following steps:

step 1, sampling an original BCG waveform by adopting a preset sampling frequency to obtain the original BCG sampling waveform, wherein the preset sampling frequency is 50;

step 2, carrying out first low-pass filtering on the original BCG sampling waveform to obtain a waveform A, and carrying out smooth filtering on the original BCG sampling waveform to obtain a waveform B;

step 2, subtracting the waveform B from the waveform A to obtain a heartbeat signal waveform C;

step 3, searching all wave crests and wave troughs in the waveform C;

step 4, counting the number of sampling points between all wave crests and wave crests in the waveform C within a preset time, and counting the number of sampling points between all wave troughs and wave troughs in the preset time, wherein the preset time is 20 seconds;

step 5, obtaining a maximum peak value according to the number of the counted sampling points between the wave crests, wherein the maximum peak value is a value with the largest number of sampling points between the wave crests, and calculating the heart rate according to the formula (1):

heart rate 60 sample frequency/max peak (1)

The sampling frequency in equation (1) is a predetermined sampling frequency.

Step 5 also includes judging the heartbeat quality, which is as follows:

acquiring a second large peak value and a third large peak value, wherein the second large peak value is a numerical value with a second maximum of numerical values of sampling points between the wave crests, the third large peak value is a numerical value with a third maximum of numerical values of the sampling points between the wave crests, and calculating the second large peak value-the maximum peak value and the third large peak value-the second large peak value when:

and when the maximum peak value is the second large peak value, and the maximum peak value is the third large peak value, and the second large peak value, the heartbeat quality is judged to be good.

Step 6, carrying out second low-pass filtering on the waveform A to obtain a first respiratory waveform, removing a direct-current component from the waveform A to obtain a second respiratory waveform, normalizing the second respiratory waveform, and multiplying the second respiratory waveform by a preset constant, wherein the preset constant is 200 to obtain a third respiratory waveform;

and 7, searching the wave crest of the second respiratory waveform, and calculating the variation coefficient by using the size and the interval of the wave crest.

And 8, calculating the multiples of the second respiratory waveform and the third respiratory waveform, and when the coefficient of variation is less than 0.2, adjusting the amplification factor of the circuit for acquiring the original BCG signal to enable the multiples of the second respiratory waveform and the third respiratory waveform to be 0.5-2.0.

And 9, detecting the variance of the third respiratory waveform of the last time period by adopting a variance method, wherein the last time period is 5 seconds, when the variance is smaller than a threshold value, the apnea is set, and when the apnea continuously reaches a preset pause time, the preset pause time is 20 seconds, a wake-up signal is sent. The wake-up signal is a voice signal.

And step 10, after the awakening signal is sent out, stopping the awakening signal when the apnea is not detected.

The method comprises the following specific steps:

1. the original BCG waveform is collected by using the sampling rate of the single chip microcomputer 50 HZ.

2. And (4) performing low pass on the original waveform, and filtering ripple interference to obtain a waveform I.

3. And obtaining a second waveform for filtering most of the heartbeat peaks by utilizing smooth filtering.

N: sliding window length, taking 12 at 50HZ samples, a: waveform two after sliding, x: low-pass filtering BCG signal without ripple wave

4. And subtracting the waveform second sequence from the waveform first sequence to obtain a waveform third, namely the heartbeat signal.

d[i]＝x[i]-a[i]

d, obtaining a heartbeat signal, x, BCG signal, a, waveform two, i: sequence of

5. And searching all wave crests and wave troughs of the waveform III to form a wave crest first array and a wave trough second array.

6. And searching the wave crest in the wave crest first array to obtain a wave crest two. And obtaining a second trough in the same way.

7. At the moment, peak values belonging to heartbeat exist in the second wave crest and the second wave trough, the number of interval points between the wave crest and the wave crest in all 20 seconds is counted, and the number of interval points between the wave trough and the wave trough in all 20 seconds is counted.

8. The maximum peak value a, the second large peak value b and the third large peak value c are obtained by using the result of the statistical peak value interval distance,

if the condition that a is equal to b-a is equal to c-b exists, the heart rate waveform has better quality. At this time:

heart rate 60/a;

9. and low-pass filtering the waveform one to obtain a respiratory waveform one.

10. And removing the direct current component of the respiratory waveform I to obtain a respiratory waveform II.

11. And normalizing the respiratory waveform II and then multiplying the normalized respiratory waveform II by 200 to obtain a respiratory waveform III.

12. And searching the wave crest of the second waveform, and calculating the variation coefficient by using the size and the interval of the wave crest.

13. And calculating the multiples of the second amplitude of the waveform and the third amplitude of the waveform. When the coefficient of variation is smaller, the single chip sends an instruction to adjust the amplification factor of the hardware circuit, so that the factor of the second waveform and the factor of the third waveform are equal to 1.

14. And detecting the variance of the waveform of the last 5 seconds by using a variance method, and detecting the apnea when the variance is smaller than a threshold value.

15. And continuously detecting the accumulated apnea for 20 seconds, and starting voice intervention to wake up the patient when the breath recovery is not detected.

16. And in the case of apnea intervention, the voice intervention is stopped when respiratory recovery and body movement are detected.

The present invention is capable of other embodiments, and various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention.