TWI790472B - Arteriosclerosis Risk Assessment System - Google Patents

Abstract

The present invention provides an arteriosclerosis risk assessment system, comprising: a wearable detection device, used to detect a heart rate signal of a subject; a motion monitoring device, wirelessly connected to the wearable detection device, and the motion detection device is used to receive the Heart rate signal, and calculate according to the heart rate signal to obtain an arteriosclerosis index, a heart rate variation index and the ratio of the heavy sine wave, and compare the upper limit value of the arteriosclerosis index, the threshold value of the heart rate variation index and the ratio of the heavy sine wave respectively Limit value, and then converted into a risk index according to the result after comparison; a cloud server, which is wirelessly connected to the mobile monitoring device, and is used to receive the arteriosclerosis index of multiple subjects and calculate the upper limit of the arteriosclerosis index value, thereby providing users with the ability to easily and quickly assess the risk of arteriosclerosis.

Description

Arteriosclerosis Risk Assessment System

An arteriosclerosis risk assessment system, especially an analysis and evaluation system that converts the heart rate signal from the time domain to the frequency domain, and can eliminate the heavy sine wave and leave the dicrotic wave to calculate the arteriosclerosis index.

Arteriosclerosis is a broad term, including small arteries, arterial media, atherosclerosis (Atherosclerosis), the most serious is that atherosclerosis occurs in coronary arteries and middle arteries, if there is no regular cardiovascular examination , The patient may not know that he is suffering from arteriosclerosis, which eventually leads to sudden myocardial infarction and stroke.

Therefore, an existing measuring device is a wearable device. Common wearable devices are mostly worn on the wrists of the subjects in the form of bracelets or watches, such as smart wearable devices such as Xiaomi bracelets and Apple Watches. The wearable device can detect a heart rate signal (generally referred to as PPG signal) of the subject through photoplethysmography (Photoplethysmography, hereinafter referred to as PPG). The information presented by the heart rate signal includes heartbeat and blood flow inside the blood vessel. state. With the PPG measurement technology, the heartbeat and blood flow status of the subject can be quickly obtained, and the subject's arteriosclerosis index (SI) can be easily measured to determine the degree of hardening of the subject's blood vessels. The arteriosclerosis index The calculation formula is:

SI=h/ΔT*100%; where h is the height of the subject, and ΔT is the time difference between the main pulse peak and the dicrotic wave in the PPG signal of the subject.

Please refer to Fig. 15, there are two pulse waves in the PPG signal of a heartbeat cycle, one is the main peak MP of the pulse wave, and the other is the dicrotic wave 40. (rebound) to the aortic valve and then transmitted to the sensing site, so the amplitude of the dicrotic wave 40 can represent the expansion ability of the arterial diameter. Once the dicrotic wave 40 is detected, the time difference ΔT between the main pulse peak MP and the dicrotic wave 40 can be further measured, and the atherosclerosis index can be calculated according to the above formula.

Please refer to FIG. 16A , but in the actual measurement, it is often found that the dicrotic wave of the subject's PPG signal is not obvious. This waveform is called a heavy sine wave 50 , and FIG. Please refer to Fig. 16B. After performing a differentiation on the heavy sine wave 50, the waveform 501 can be obtained, but there is still no band showing blood vessels rebounding blood, and the exact position of the dicrotic wave cannot be found. Instead, it is captured near the trough, and the trough occurs The time of ΔT will be later than the time of dicrotic wave, so ΔT will become larger, resulting in low arteriosclerosis index. The reason for the heavy sine wave 50 is that the subject’s blood vessels are aging, hardened, or due to poor blood circulation. After the heart is compressed to pump blood, although the blood flows into the blood vessels, the blood vessels lose their ability to expand due to hardening, and cannot rebound blood, resulting in inability to Creates a tremor wave. According to statistics, more than 50% of older people with a history of cardiovascular disease have a heavy sine wave 50 in their PPG signal, and these patients are the most important group to be tracked by the atherosclerosis index. In addition, there are also technologies on the market that claim to be able to monitor arteriosclerosis, mainly through full-time monitoring of HRV performance to estimate cardiovascular status. The reference indicators for HRV are time-domain indicators such as SDNN, SDANN, SDNNIndex, etc. Accurate judgment of indicators requires PPG signal 24 hours a day for statistics, and the interference of irregular heart rate caused by exercise or overactive interactive sensory nerve due to a large number of activities must be excluded. It is only suitable for patients who are bedridden in the hospital. use. Therefore, how to allow individuals to easily assess their own possibility of suffering from arteriosclerosis in daily life has become an important goal in cardiovascular technology.

In order to allow users to easily assess the risk of suffering from arteriosclerosis in daily life, the present invention proposes an arteriosclerosis risk assessment system, which measures heart rate signals by wearing a detection device, and converts the heart rate signal into arteriosclerosis by a mobile monitoring device Relevant data provide users with self-detection to obtain instant and rapid detection results of arteriosclerosis risk.

In order to achieve the above object, the arteriosclerosis risk assessment system of the present invention includes: A wearing detection device, which is used to detect a heart rate signal of a subject; An action monitoring device, which is wirelessly connected to the wearable detection device, the action monitoring device is used to receive the heart rate signal, and calculate an arteriosclerosis index, a heart rate variation index and the ratio of the heavy sine wave according to the heart rate signal, and The arteriosclerosis index, the heart rate variability index and the ratio of the heavy sine wave are respectively compared with an upper limit value of the arteriosclerosis index, a threshold value of the heart rate variation index and a upper limit value of the heavy sine wave ratio, and then converted into a risk index according to the comparison result ; A cloud server, which is wirelessly connected to the action monitoring device, and is used to receive the arteriosclerosis index of each of the plurality of subjects, and calculate the upper limit of the arteriosclerosis index.

After the heart rate signal obtained by the photoplethysmography method, the present invention further calculates the ratio of the arteriosclerosis index, the heart rate variation index and the heavy sine wave according to the heart rate signal, and compares an arteriosclerosis index upper limit, a heart rate The threshold value of the variation index and the proportional risk value of a double sine wave; according to the comparison results, it can provide users with a risk reference for the possibility of arteriosclerosis in their daily lives, so as to be able to prevent accidents before they occur. Seek medical attention early.

Furthermore, the present invention can further eliminate the heavy sine wave in the heart rate signal, leaving the dicrotic wave capable of calculating the arteriosclerosis index, that is, it can eliminate the error calculated by using the heavy sine wave, and use the pure dicrotic wave to calculate the The arteriosclerosis index can obtain a more accurate arteriosclerosis index. In the present invention, the average value of the arteriosclerosis index of the subject can be obtained by summing up multiple arteriosclerosis indices, so as to prevent a single arteriosclerosis index from being affected by measurement errors and improve measurement accuracy.

The present invention discloses an arteriosclerosis risk assessment system, which can eliminate heavy sine waves in heart rate signals, leaving dicrotic waves for calculating the arteriosclerosis index, so as to improve the accuracy of calculating the arteriosclerosis index.

Please refer to FIG. 1 , the arteriosclerosis risk assessment system of the present invention includes: a wearing detection device 10 , an activity monitoring device 20 and a cloud server 30 .

The wearing detection device 10 can be worn on a subject to detect a heart rate signal of the subject, wherein the heart rate signal is presented as an amplitude-time waveform in the time domain. Generally speaking, the wearing detection device 10 can be a smart bracelet, and the subject can wear the smart bracelet on his wrist. The wearing detection device 10 obtains the heart rate signal through photoplethysmography (hereinafter referred to as PPG), and the heart rate signal may include the heartbeat and the blood flow inside the blood vessel. Specifically, the wearing detection unit 10 may include a control unit 11 , a light emitting unit 13 and a light receiving unit 15 , and the light emitting unit 13 and the light receiving unit 15 are respectively electrically connected to the wearing detection unit 10 . The control unit 11 is used to activate the light emitting unit 13 and the light receiving unit 15, and has a wireless transmission function; the light emitting unit 13 is used to emit a continuous detection light to the blood in the blood vessel of the subject, The blood in the blood vessel reflects the detection light to form a reflected light; the light receiving unit 15 receives the reflected light and transmits it to the control unit 11, and the control unit 11 converts the reflected light into the heart rate signal. Since the light-emitting unit 13 continuously emits the detection light to the blood in the blood vessel, the reflected light is also continuous, and the control unit 11 converts the continuous reflected light into amplitude (intensity of reflected light) and time relationship, which is the heart rate signal.

The activity monitoring device 20 is wirelessly connected to the wearing detection device 10 and is also held by the subject. The activity monitoring device 20 can be a mobile phone or a tablet, and can be connected with the wearing detection device 10 via Bluetooth. The activity monitoring device 20 is used to receive the heart rate signal, and calculate an arteriosclerosis index, a heart rate variation index and the ratio of the heavy sine wave according to the heart rate signal, and calculate the arteriosclerosis index, the heart rate variation index and the heavy sine wave The ratio of an arteriosclerosis index upper limit value, a heart rate variability index threshold value and a double sine wave ratio upper limit value are compared respectively, and then converted into a risk index according to the comparison result, wherein the risk index represents that the subject suffers from arteriosclerosis-related disease risk. The action monitoring device 20 can further analyze and monitor the subject's pulse, blood pressure, and heart rate variability indicators (such as TP, HF, and LF) for a long period of time according to the heart rate signal, and provide the monitored pulse, blood pressure, and heart rate variability indicators. To the subject, let the subject evaluate his own health.

The cloud server 30 is wirelessly connected to the activity monitoring device 20 to receive and store the arteriosclerosis index, the user's pulse, blood pressure, and heart rate variability index.

Among them, the cloud server 30 can also collect the latest confirmed number of people suffering from arteriosclerosis-related diseases provided by credible health research institutions such as the Ministry of Health and Welfare, and compare it with the total population of the year ( Multiple subjects), calculate the proportion of subjects who do not suffer from arteriosclerosis-related diseases, use the proportion of subjects who do not suffer from arteriosclerosis as the threshold of arteriosclerosis, and then use the collected arteriosclerosis index data to linearly The method of regression calculation is used to obtain the upper limit value of the arteriosclerosis index, the threshold value of the heart rate variation index and the upper limit value of the heavy sine wave ratio of the multiple subjects, and the range of the arteriosclerosis index higher than the upper limit value is regarded as the suffering High-risk group indicators for arteriosclerosis-related diseases. Among the multiple subjects, it is possible to further calculate the proportion of men and women in each age group who do not suffer from arteriosclerosis-related diseases, and use the proportion of men and women who do not suffer from arteriosclerosis as the threshold of arteriosclerosis, and then through the collected arteriosclerosis index data In the way of linear regression calculation, the upper limit value of the arteriosclerosis index, the threshold value of the heart rate variation index and the upper limit value of the heavy sine wave ratio of the plurality of subjects of each age group and sex are obtained, and the arteriosclerosis index A range higher than the upper limit is used as an indicator of a high-risk group suffering from arteriosclerosis-related diseases. The formula for linear regression is:

，其中，

；

,in,

;

Although those who exceed the upper limit may belong to the high-risk group of suffering from arteriosclerosis, but it is still impossible to know the location of the arteriosclerosis. In this regard, heart rate variability related experiments and studies have pointed out that for patients with coronary artery disease, the total spectral power (TP) in the heart rate variability (HRV) is much smaller than that of normal people, so the reference value of whether the heart rate variability is abnormal , is to use the recommended reference value of heart rate variability (as shown in Table 1), and subtract three standard deviations from the total spectral power (TP) in the recommended normal range to obtain a threshold of heart rate variability lower than 99.85% of the public. In Table 1, 3466-1018*3=412, no matter how tired a healthy person is, the total spectral power (TP) cannot be lower than 412, so 412 is regarded as the lower limit threshold, and if it is lower than 412, it means that the body has Disease arises.

3466±1018*3 LF

1170±416 HF

975±203 表1、心率變異頻域指數正常範圍 parameter unit The normal range obtained by the current experiment TP

3466±1018*3 LF

1170±416 HF

975±203 Table 1. Normal range of heart rate variability frequency domain index

Use the two thresholds of TP value and SI value to establish high-risk groups, medium-risk groups, and low-risk groups. For example, a person is 36 years old, and the SI value is 9.7 (the average value of all SI of the subject) greater than the confidence interval. If the TP value is also less than the lower limit of heart rate variability, it is judged as a high-risk group; if the TP value falls within the normal range, it is judged as a medium-risk group and informed of the possibility of arteriosclerosis. A more in-depth coronary artery examination should be carried out as soon as possible, and the judgment principles are shown in Table 2.

risk group condition high risk 1. SI value is higher than the upper limit value 2. TP is lower than the set lower limit value medium risk 1. SI value is higher than the upper limit 2. TP falls within the normal range low risk 1. SI value is within normal range or lower than normal range 2. TP falls within normal range Table 2. Risk group comparison table

According to the judging principle of Table 2, the present invention can be further applied to the patients of all ages and genders who have been diagnosed with cardiovascular disease to measure PPG and count the occurrence ratio of heavy sine waves, so as to determine the high risk ratio (such as 55%) ), as a reference for further judging whether the subject is a high-risk group suffering from arteriosclerosis-related diseases. Finally, according to the proportion of SI exceeding the upper limit, TP value and the proportion of heavy sine waves within the test time (for example, 100 seconds), determine the risk level of the subject suffering from arteriosclerosis-related diseases, as shown in Table 3, the higher the risk index It means that the subject has a higher risk of suffering from arteriosclerosis-related diseases.

SI distribution ratio of subjects within the test time Is TP below the normal range heavy sine wave ratio risk index All SIs are below the upper limit no 0% 0 All SIs are below the upper limit yes 0% 0 but recommends immediate rest More than 50% of the SI distribution is below the upper limit no lower than high risk ratio 0 More than 50% of the SI distribution is below the upper limit yes lower than high risk ratio 1 (It is recommended to measure again after a full rest) More than 50% of the SI distribution is below the upper limit no higher than high risk ratio 2 More than 50% of the SI distribution is below the upper limit yes higher than high risk ratio 3 (It is recommended to measure again after a full rest) More than 50% of the SI distribution is above the upper limit no lower than high risk ratio 3 More than 50% of the SI distribution is above the upper limit yes lower than high risk ratio 4 (It is recommended to measure again after a full rest) More than 50% of the SI distribution is above the upper limit no higher than high risk ratio 4 More than 50% of the SI distribution is above the upper limit yes higher than high risk ratio 5 undetectable 100% 5 Table 3. Reference table of degree of arteriosclerosis

In terms of application, the wearing detection device 10 is responsible for collecting relevant personal physiological information in real time, and storing the physiological information in the subject's action monitoring device 20, and using the latest big data analysis results of the cloud server for analysis and diagnosis. At the same time, personal physiological information is regularly provided to the cloud server 30 for big data analysis. Therefore, the cloud server 30 can not only collect big data through a large number of subjects' action monitoring devices 20, but also receive statistical data from various regional health care units at any time, and provide the latest and complete reference indicators of the subjects at any time.

In practical applications, the action monitoring device 20 can not only receive personal physiological data from the wearable detection device 10, but also establish the monitoring of the physical condition of the subject, the risk of cardiovascular disease, long-term depression/manic depression/hypertension, etc. through calculation and processing. In addition to the analysis of labor and rehabilitation conditions, the update information of the cloud server 30 can be used to update the parameters of various judgment functions in real time according to the increasingly complete big data data, resulting in more accurate monitoring and judgment.

Based on the atherosclerosis index analysis system described above, the atherosclerosis index analysis method of the present invention will be described below with reference to FIG. 2 to FIG. 5 .

The method of the present invention comprises the following steps:

S11: Divide the heart rate signal HR into continuous complex sub-heart rate signals HR1, and generate a frequency pulse signal HF1 through Fourier transform for each sub-heart rate signal HR1;

S12: Calculate an S value of each sub-heart rate signal HR1, the S value represents:

；

;

Wherein A _first-f is the intensity value of the frequency pulse signal HF1 at one frequency, and A _second-f is the intensity value of the frequency pulse signal HF1 at twice the frequency; if the S value is greater than or equal to a first threshold value, then remove The sub-heart rate signal HR1; if the S value is less than the first threshold, perform the next step;

S13: Comparing a time ratio with a second threshold; wherein, calculating a first time difference ΔT ₁ and a second time difference ΔT ₂ of the sub-heart rate signal HR1, and dividing the first time difference ΔT ₁ by the second time difference ΔT ₂ Obtain the time ratio, wherein the first time difference ΔT ₁ represents the time difference between a main heart rate peak MP1 and a heart rate secondary peak SP1 of the sub-heart rate signal HR1, and the second time difference ΔT ₂ represents the main heart rate peak of the sub-heart rate signal HR1 The time difference between peak MP1 and a main valley MH1; if the time ratio is less than a second threshold, execute the next step; if the time ratio is greater than or equal to a second threshold, then return to step S11; calculate the first time difference ΔT ₁ and The method of the second time difference ΔT ₂ can generate a differential signal by differentiating the sub-heart rate signal once, and find the place where the slope of the differential signal is 0, which is the main peak MP1 of the heart rate, the sub-peak SP1 of the heart rate and the main valley MH1 Calculate the time elapsed from the main heart rate peak MP1 to the heart rate sub-peak SP1 as the first time difference ΔT ₁ , and calculate the time elapsed from the heart rate sub-peak SP1 to the main trough MH1 as the second time difference ΔT ₂ .

S14: Calculate the average value of an arteriosclerosis index of the subject; if the time ratio is less than the second threshold, calculate the arteriosclerosis index (SI) according to the subject's height and the first time difference _ΔT1 , and calculate After multiple entries of the arteriosclerosis index are obtained, an average value of the arteriosclerosis index is obtained from the plurality of entries of the arteriosclerosis index, wherein the calculation formula of the arteriosclerosis index is:

，其中h為該受測者的身高；

, where h is the height of the subject;

S15: Calculate the ratio of the heavy sine wave; divide the number of the sub-heart rate signals HR1 whose attribute is the heavy sine wave by the total number of all the sub-heart rate signals to obtain the ratio of the heavy sine wave.

Please refer to FIG. 3A, in step S11, the wearing detection device 10 detects the subject and sends the heart rate signal HR to the activity monitoring device 20, and the activity monitoring device 20 distinguishes the heart rate signal HR into continuous complex numbers The sub-heart rate signal HR1, that is to say, the heart rate signal HR is composed of the complex sub-heart rate signals HR1. Please further refer to FIG. 3B , the activity monitoring device 20 performs a Fourier transform on each sub-heart rate signal HR1 to obtain the frequency pulse signal HF1 , in other words, each sub-heart rate signal HR1 corresponds to a frequency pulse signal HF1 . In this step, the sub-heart rate signal HR1 belonging to the time domain is transformed through Fourier to generate the frequency pulse signal HF1 belonging to the frequency domain. It can be seen from FIG. 3B that the frequency pulse signal HF1 has two peaks (intensities), one of which is located at the occurrence of one frequency, and the other peak is located at the occurrence of double the frequency.

In step S12, the activity monitoring device 20 takes out the intensity value of the frequency pulse signal HF1 at one frequency and the intensity value at twice the frequency, and calculates the intensity value of the pulse signal HF1 at one frequency and the intensity value at two frequencies. The ratio between the intensity values of the doubling frequency yields the S value. The activity monitoring device 20 then calculates the magnitude relationship between the S value and the first threshold value. When the S value is greater than or equal to the first threshold value, the activity monitoring device 20 discards the sub-heart rate signal HR1 corresponding to the S value; If the value is smaller than the first threshold, the activity monitoring device 20 retains the sub-heart rate signal HR1 corresponding to the S value and executes the next step. The reason for discarding the S value greater than or equal to the first threshold is that the S value greater than or equal to the first threshold represents that the intensity value of the pulse signal HF1 at one frequency is too different from the intensity value at twice the frequency, indicating that The frequency pulse signal HF1 belongs to the heavy sine wave, and the heavy sine wave cannot be used to calculate the arteriosclerosis index (SI), so discarding the frequency pulse signal HF1 belonging to the heavy sine wave helps to improve the calculation of the arteriosclerosis index (SI). ) accuracy.

The following uses the actual operation as an example to introduce. Please refer to FIG. 3A and FIG. 3B. In FIG. 3A, the heart rate signal HR obtained by the wearable detection device 10 using PPG technology is presented. The detection object is a subject with normal blood vessel function, so it can be seen in FIG. 3A The heart rate signal is mostly composed of dicrotic waves. Fig. 3B is a frequency domain diagram of the frequency pulse signal HF1 generated after performing step S11 on one of the sub-heart rate signal HR1. It can be seen that the double frequency is 1.75 Hz, and the peak strength (A _first-f ) at the double frequency is 2168; the double frequency falls roughly at 3.375Hz, the peak intensity of the double frequency (A _second-f ) is 1590, the first threshold is preset at 3.5, the S value is 2168/1590=1.36, and 1.36<3.5, This means that there is little difference between the intensity value of one frequency and the intensity value of twice the frequency. The frequency pulse signal HF1 has two peaks of obvious intensity, which means that the sub-heart rate signal HR1 may be a dicrotic wave, so the heart rate signal HR has a certain degree of The reference value can proceed to the next step.

Please refer to FIG. 4A and FIG. 4B. In FIG. 4A, the heart rate signal HR obtained by the wearable detection device 10 using PPG technology is also presented. The detection object is a subject with abnormal blood vessel function, so it can be seen in FIG. 4A The heart rate signal HR is mostly composed of sub-heart rate signals HR2 belonging to heavy sine waves, and FIG. 4B is a frequency-domain diagram of the pulse frequency signal HF2 generated after performing step S11 on one of the sub-heart rate signals HR2. As can be seen in Figure 4B, the double frequency is 1.25Hz, and the intensity (A _first-f ) at the double frequency is 3271; the double frequency falls roughly at 2.5Hz, and the strength of the double frequency (A _second-f ) is 791.6, the first threshold is preset as 3.5, and the S value is 3271/791.6=4.13≧3.5, which means that the difference between the intensity value of one frequency and the intensity value of two times is too large, and the pulse signal HF2 has only one obvious The peak of the intensity indicates that the sub-heart rate signal HR2 is a heavy sine wave, so the sub-heart rate signal HR2 has no reference value, and the activity monitoring device 20 discards the sub-heart rate signal HR2.

Setting the first threshold to 3.5 is a better value based on the results of multiple experiments. On the premise that the first threshold is preset to 3.5, the sub-heart rate of the heavy sinusoid and dicrotic wave can be screened out more accurately. signals HR1 and HR2 , so in practical applications, the first threshold is preferably preset to be 3.5.

Although in step S12, the sub-heart rate signal HR2 which is a heavy sine wave is found and discarded, the undiscarded complex sub-heart rate signal HR1 may also be a relatively inconspicuous heavy sine wave, so the undiscarded complex sub-heart rate signal HR1 must be taken out The complex sub-heart rate signal HR1 executes step S13 to more accurately screen out the sub-heart rate signal HR1 that is actually a dicrotic wave.

Please refer to FIG. 5 , in step S13 , the activity monitoring device 20 differentiates each sub-heart rate signal HR1 retained in step S12 , so each sub-heart rate signal HR1 generates a differential signal. In terms of waveform, the sub-heart rate signal HR1 has the main heart rate peak MP1 , the heart rate sub-peak SP1 and the main trough MH1 . The activity monitoring device 20 calculates the first time difference ΔT ₁ from the main heart rate peak MP1 to the heart rate sub-peak SP1 , and the second time difference ΔT ₂ from the heart rate main peak MP1 to the main heart rate trough MH1 . The activity monitoring device 20 calculates the ratio of the first time difference ΔT ₁ and the second time difference ΔT ₂ to obtain the time ratio, and compares the time ratio with the second threshold, if the time ratio is smaller than the second threshold, it means In the sub-heart rate signal HR1, the first time difference _ΔT1 between the main heart rate peak MP1 and the heart rate sub-peak SP1 is longer, which means that the sub-heart rate signal HR1 has more obvious heart rate main peak MP1 and the heart rate sub-peak. peak SP1, it can be determined that the sub-heart rate signal HR1 belongs to the dicrotic wave, and the activity monitoring device 20 retains the sub-heart rate signal HR1; on the contrary, if the time ratio is greater than or equal to the second threshold, it means that the sub-heart rate signal HR1 , the first time difference _ΔT1 between the heart rate main peak MP1 and the heart rate secondary peak SP1 is relatively short, which is equivalent to that the sub-heart rate signal HR1 has only one obvious peak (it cannot be distinguished whether the peak is the heart rate main peak MP1 or the heart rate sub-peak SP1), it can be determined that the sub-heart rate signal HR1 belongs to a heavy sine wave, and the activity monitoring device 20 discards the sub-heart rate signal HR1. After repeated experiments and calculations, it is found that when the second threshold is 80%, it is more accurate to distinguish dicrotic waves and baric sinusoids, so in practical applications, the second threshold is preset to 80%.

In step S14, the activity monitoring device 20 takes out the first time difference ΔT ₁ of the retained sub-heart rate signal HR1, and substitutes the first time difference ΔT ₁ into the calculation formula of the arteriosclerosis index and calculates a plurality of the arteriosclerosis By taking the average of the indices, the average value of the arteriosclerosis index of the subject can be obtained. For example, in step S11, the wearable detection device 10 divides the heart rate signal HR into 10 sub-heart rate signals HR1 and HR2, and in step 12 eliminates a sub-heart rate signal HR2 whose attribute is a heavy sine wave. In S13, another sub-heart rate signal HR2 whose attribute is a heavy sine wave is eliminated, so eight sub-heart rate signals HR1 are retained. The activity monitoring device 20 extracts the respective first time differences ΔT ₁ of the eight sub-heart rate signals HR1 and substitutes them into the calculation formula of the arteriosclerosis index to obtain eight arteriosclerosis indices. The activity monitoring device 20 calculates the average value of the eight arteriosclerosis indices, that is, obtains the average value of the arteriosclerosis index of the subject.

)。In step S15, since the sub-heart rate signal HR1 belonging to the dicrotic wave and the sub-heart rate signal HR2 belonging to the heavy sine wave are distinguished in steps S12 and S13, the activity monitoring device 20 further classifies the sub-heart rate signal HR2 belonging to the heavy sine wave The number of sub-heart rate signals HR2 is compared with the total number of all sub-heart rate signals HR1, HR2 to obtain the ratio of the heavy sine wave. For example, in step S11, the wearable detection device 10 divides the heart rate signal into ten sub-heart rate signals HR1 and HR2, and eliminates a sub-heart rate signal HR2 whose attribute is a heavy sine wave in step S12, and in step S13 Eliminate another sub-heart rate signal HR2 whose attribute is a heavy sine wave, so there are two sub-heart rate signals HR2 that belong to a heavy sine wave, and the proportion of the heavy sine wave can be obtained as 20% (

).

Utilizing the above steps, the present invention can filter the heart rate signal HR, identify and exclude the wave band of the dicrotic wave, leave the wave band of the dicrotic wave, and use the first time difference ΔT ₁ of the dicrotic wave to calculate the average value of the arteriosclerosis index , to prevent incorrect calculation of the average value of the arteriosclerosis index due to heavy sine waves. In step S12, first convert each sub-heart rate signal HR1 from the time domain to the frequency domain, and calculate the S value of each sub-heart rate signal, and then evaluate the size of the S value and the first threshold, and eliminate the strength that is too high The S value of each sub-heart rate signal HR1 is differentiated once to obtain the dicrotic wave, and the first time difference ΔT ₁ between the heart rate main peak MP1 and the heart rate sub-peak SP1 in the dicrotic wave is calculated, and the The second time difference ΔT ₂ between the heart rate main wave peak MP1 and the main wave trough MH1 is compared to filter out the heavy sine wave waveform again, so that the heavy sine wave can be effectively found and discarded, and the heavy beat wave can be retained. Bobo calculates multiple more accurate arteriosclerosis indices, and averages the sum of multiple arteriosclerosis indices to obtain the correct average value of the arteriosclerosis index.

In addition to the above-mentioned arteriosclerosis risk assessment system, the arteriosclerosis index analysis system of the present invention can further perform another measurement step of the arteriosclerosis index that removes the heavy sine wave. After the arteriosclerosis risk assessment system is implemented, further execution Said steps can assist in removing the heavy sine wave bands to ensure that the remaining bands are all dicrotic waves, which is more accurate when calculating the arteriosclerosis index (SI) and the average value of the arteriosclerosis index.

Please refer to Fig. 6 to Fig. 14B, another kind of arteriosclerosis index measurement method that removes heavy sine wave of the present invention comprises the following steps:

Please refer to FIG. 6 and FIG. 7A, S21: establish an auxiliary line SL of the sub-heart rate signal HR1, and the auxiliary line SL extends linearly from a main heart rate peak MP2 of the sub-heart rate signal HR1 to a main heart rate valley MH2; wherein the heart rate The main peak MP2 represents the maximum amplitude of the sub-heart rate signal HR1, and the main heart rate valley MH2 represents the lowest amplitude of the sub-heart rate signal HR1;

S22: Establish an auxiliary line distance signal SLS and calculate the number of troughs of the auxiliary line distance signal SLS; please refer to FIG. 7B , the activity monitoring device 20 differentiates the sub-heart rate signal HR1 once to obtain the differential signal HD1, and obtains the differential signal HD1 A first time point t ₁ and a second time point t ₂ , and calculate the distance between the auxiliary line SL and the _sub -heart rate signal HR1 between the first time point t 1 and the second time point t ₂ ; The first time point _t1 is the occurrence time of the first valley of the differential signal HD1 along the positive direction of the time axis, and the second time point _t2 is the occurrence time of the heart rate main valley MH2. Please refer to Fig. 7C, the auxiliary line distance signal SLS represents all vertical distances between the auxiliary line SL and the sub-heart rate signal HR1 in the interval from the first time point _t1 to the second time point _t2 distance function.

Fig. 7A to Fig. 7C present the process of actually performing steps S21 and S22 on the sub-heart rate signal HR1 belonging to the dicrotic wave. In Fig. 7C, it can be found that the auxiliary line distance signal SLS has two obvious valley points H1 and H2, It means that the sub-heart rate signal HR1 may belong to the dicrotic wave.

In the following, the steps S21 and S22 are performed using the sub-heart rate signal HR2 of the heavy sine wave as a practical example. Please refer to FIG. 8A , in step S21 , the auxiliary line SL of the sub-heart rate signal HR2 is firstly established, and the auxiliary line SL extends linearly from the main peak MP2 of the sub-heart rate signal HR2 to the main valley MH2 of the heart rate. Please refer to Fig. 8B, in step S22, differentiate the sub-heart rate signal HR2 once to obtain the differential signal HD2, take a first time point t ₁ and a second time point t ₂ of the differential signal HD2, and calculate The distance between the auxiliary line SL and the sub-heart rate signal HR2 between the first time point _t1 and the second time point _t2 . It can be found in FIG. 8C that the auxiliary line distance signal SLS has only one obvious trough, which means that the sub-heart rate signal HR2 belongs to a heavy sine wave.

Although in steps S21 and S22 it can be determined that the sub-heart rate signals HR1 and HR2 belong to dicrotic waves or heavy sinusoidal waves, when actually measuring the waveforms of the sub-heart rate signals HR1 and HR2, it will be found that the sub-heart rate signals HR1 and HR2 may It is a heavy sine wave or an inconspicuous dicrotic wave (a waveform similar to a heavy sine wave but actually a dicrotic wave). Taking the actual sub-heart rate signal as an example, please refer to FIG. 9A. After the sub-heart rate signal HR3 establishes the auxiliary line SL, the auxiliary line distance signal SLS is obtained as shown in FIG. 9B. It can be found that the auxiliary line distance signal SLS has An obvious valley point H4, and an inflection point H5, whether the inflection point H5 is a valley point will affect the judgment of which waveform the sub-heart rate signal HR3 belongs to, so it is difficult to distinguish whether the sub-heart rate signal HR3 belongs to the dicrotic wave or is a heavy sine wave. Therefore, it is necessary to further quote the calculation formula of the Bezier curve, and adopt different judgment and calculation methods for the dicrotic wave, the inconspicuous dicrotic wave and the dicrotic wave. In the following introduction, step S23 is a method for calculating Bezier curves, step S231 is a calculation process for dicrotic waves, and step S232 is a calculation process for inconspicuous dicrotic waves and dicrosine waves.

Please refer to Fig. 10A to Fig. 10C, S23: calculate a Bezier curve, the calculation formula of the Bezier curve is:

；

;

And take the first time point t ₁ , that is, the first trough occurrence time (t ₁ ) of the auxiliary line distance signal SLS along the time axis to obtain B(t ₁ ), and take the second time point t ₂ , that is, B(t ₂ ) is obtained from the second trough occurrence time (t ₂ ) of the auxiliary line distance signal SLS along the positive time axis.

；in,

;

Please refer to Fig. 10B, P ₀ is the first valley point along the positive direction of the time axis after the sub-heart rate signal HR1 is differentiated once; please refer to Fig. 10A, P ₃ is the amplitude at a third time point _t3 , the third Time point _t3 is the time when the heart rate main trough occurs; _P1 is a first control point, and the first trough occurrence time (t ₁ ) of the auxiliary line distance signal SLS along the positive time axis is substituted into the Bezier curve obtained from the calculation formula (that is, B(t ₁ )); P ₂ is a second control point, which is substituted by the second trough occurrence time (t ₂ ) of the auxiliary line distance signal SLS along the positive direction of the time axis into the It is obtained from the calculation formula of the Purcell curve (that is, B(t ₂ )).

S231: Substitute P ₀ , P ₃ , B(t ₁ ) and B(t ₂ ) into the calculation formula of the Bezier curve and calculate three times to obtain P ₁ and P ₂ , that is, according to P ₁ , P ₂ , P ₀ and P ₃ obtain the Bezier curve BC1 of the sub-heart rate signal HR1 (shown in FIG. 10A ). This method is applicable to the sub-heart rate signal HR1 (shown in FIG. 10C ) where the auxiliary line distance signal SLS has two obvious troughs, which means that the sub-heart rate signal HR1 is a dicrotic wave.

Please refer to Fig. 11A to Fig. 11C, S232: Substitute P ₀ , P ₃ , B(t ₁ ), B(t ₂ ) into the calculation formula of the Bezier curve and calculate three times to obtain P ₁ and P ₂ , namely The Bezier curve BC2 (shown in FIG. 11A ) of the sub-heart rate signal HR2 can be obtained according to P ₁ , P ₂ , P ₀ , and P ₃ . The calculation method of B(t ₂ ) is:

；

;

This method is applicable to the sub-heart rate signal HR2 (shown in FIG. 11C ) where the auxiliary line distance signal SLS has only one obvious trough, which means that the sub-heart rate signal HR2 is a heavy sine wave or an indistinct dicrotic wave.

S24: Comparing the correlation between the Bezier curve and the heart rate signal, if a correlation coefficient between the Bezier curve and the sub-heart rate signal HR1 is less than a third threshold, calculate the arteriosclerosis index of the sub-heart rate signal HR1 (SI); if the correlation coefficient between the Bezier curve and the sub-heart rate signal HR2 is greater than or equal to a third threshold, discard the sub-heart rate signal HR2. The third threshold may be set at 99.85%.

Among them, the calculation method of the correlation coefficient is:

；

;

為相關係數，n為取樣個數，x代表該心率信號在相同時點下的y軸座標值；y代表該貝賽爾曲線在相同時點下的y軸座標值。

is the correlation coefficient, n is the number of samples, x represents the y-axis coordinate value of the heart rate signal at the same time point; y represents the y-axis coordinate value of the Bezier curve at the same time point.

Please refer to FIG. 12A. In practical applications, the solid line represents the heart rate signal HR1 actually measured by a subject, and the dotted line represents the Bezier curve BC1 derived from the heart rate signal HR1. The heart rate signal HR1 is related to the heart rate signal HR1. The correlation coefficient of the Bezier curve BC1 is 99.49%, which is less than the third threshold (99.85% by default), which means that the heart rate signal HR1 is not correlated with the Bezier curve BC1, and in FIG. 12A, the heart rate signal HR1 The range in which the amplitude is greater than the amplitude of the Bezier curve BC1 (the range between the two dotted lines) is the range in which the dicrotic wave occurs. Please refer to FIG. 12B , the heart rate sub-peak SP1 of the dicrotic wave can be found by performing a differential on the interval, which is substituted into step S13 to further calculate the arteriosclerosis index (SI). It can be seen from FIG. 12A and FIG. 12B that the position of the heart rate sub-peak SP1 indeed falls in the interval where the amplitude of the heart rate signal HR1 is greater than the amplitude of the Bezier curve BC1 .

Please refer to Fig. 13A, in another practical application, also, the solid line represents the heart rate signal HR2 actually measured by a subject, the dotted line represents the Bezier curve BC2 derived from the heart rate signal HR2, the heart rate The correlation coefficient between the signal HR2 and the Bezier curve BC2 is 99.96%, which is greater than the third threshold (99.85% by default), indicating that the heart rate signal HR2 almost overlaps with the Bezier curve BC2, which means that the heart rate signal HR2 is a heavy string Wave, therefore, the heart rate signal HR2 whose correlation coefficient with the Bezier curve BC2 is too high can be discarded. If the heart rate signal HR2 is differentiated once, the unknown point N as shown in Figure 13B will be obtained, and the unknown point N does not fall in the overlapping interval between the heart rate signal HR2 and the Bezier curve BC2 (between two dot chain lines range), but falls on the position of the non-Bezier curve, so the Bezier curve can be used to accurately distinguish whether the heart rate signals HR1, HR2 are heavy sine waves.

The reason for calculating the Bezier curve is that the Bezier curve can be regarded as the hypothetical heavy sinusoidal curve of the sub-heart rate signals HR1 and HR2, and the calculation of the correlation coefficient can further determine the relationship between the Bezier curve and the sub-heart rate signal HR1. , HR2 overlap rate (correlation level). If the overlap rate between the Bezier curve and the sub-heart rate signal HR1 is low, the correlation coefficient of the sub-heart rate signal HR1 will be smaller than the third threshold, which means that the correlation between the Bezier curve and the sub-heart rate signal HR1 is low. The heart rate signal HR1 is not exactly the same as the imaginary heavy sinusoidal curve, and the sub-heart rate signal HR1 is the dicrotic wave. If the overlap rate between the Bezier curve and the sub-heart rate signal HR2 is high, the correlation coefficient of the sub-heart rate signal HR2 will be greater than or equal to the third threshold, which means that the Bezier curve has a high correlation with the sub-heart rate signal HR2. The sub-heart rate signal HR2 is almost identical to the hypothetical heavy sine wave curve, and the sub-heart rate signal HR2 is a heavy sine wave.

Please refer to Fig. 14A, S25: Take the first interval along the positive direction of the time axis where the amplitude of the heart rate signal HR4 is greater than the amplitude of the Bezier curve BC4, and differentiate the sub-heart rate signal HR4 in the interval once, and the interval along the time The first peak in the positive direction of the axis is the differential sub-peak, and the arteriosclerosis index is calculated by taking the time difference between a heart rate sub-peak corresponding to the occurrence time of the differential sub-peak and the heart rate main peak.

Please refer to FIG. 14A and FIG. 14B , the heart rate signal HR4 includes a heart rate main peak MP4 , the heart rate secondary peak SP4 and a third heart rate peak TP4 . The pulse frequency signal HD4 is obtained after the heart rate signal HR4 is differentiated once. The pulse frequency signal HD4 includes a differential main peak MP5 , a differential secondary peak SP5 and a third differential peak TP5 . Since the differential sub-peak SP5 is closer to the differential main peak MP5 than the third differential peak TP5, the heart rate sub-peak SP4 corresponding to the occurrence time of the differential sub-peak SP5 is a real dicrotic wave. The arteriosclerosis index is calculated from the time difference (first time difference ΔT ₁ ) between the peak SP4 and the main heart rate peak MP4.

The reason for performing step S25 is that some waveforms of the heart rate signal HR4 may oscillate, and the reason may be that the measurement is affected by an external force. For example, the heart rate signal HR4 in FIG. 14A makes the Bezier curve BC4 have two heavy Pulse wave range (the range where the amplitude of the two heart rate signals HR4 is greater than the amplitude of the Bezier curve BC4), and the correlation coefficient between the heart rate signal HR4 and the Bezier curve BC4 is 99.73%, less than 99.85%, indeed It is a dicrotic wave, so the peak closest to the differential main peak MP5 must be taken, that is, the differential secondary peak SP5 is the position of the dicrotic wave, so as to avoid finding the wrong dicrotic peak due to waveform oscillation.

The present invention obtains the heart rate signal of the subject through the wearable detection device 10, and then uses the cloud data to calculate the upper limit of the arteriosclerosis index of different age groups. Just input the age and height, and the subject can be calculated. The arteriosclerosis index, the characteristic value of the heart rate variation and the ratio of the heavy sine wave of the patient are compared with the upper limit value of the arteriosclerosis index, the threshold value of the heart rate variation index and the upper limit of the heavy sine wave ratio to generate the The risk index of the subject can allow individuals to easily assess their own arteriosclerosis in daily life, so as to facilitate timely medical treatment before an emergency occurs. Furthermore, the method steps of the present invention can remove heavy sine waves, avoiding errors in the calculation of the arteriosclerosis index due to heavy sine waves generated by the human body when calculating the arteriosclerosis index, and the calculated arteriosclerosis index is more accurate.

10: Wearing detection device 11: Control unit 13: Light emitting unit 15: Light receiving unit 20: Action monitoring device 30: Cloud server 40: Stretch wave 50: Heavy sine wave 501: Primary differential heavy sine wave HR: Heart rate Signals HR1, HR2, HR3, HR4: sub-heart rate signals HF1, HF2: pulse frequency signals HD1, HD2: differential signals ΔT ₁ : first time difference ΔT ₂ : second time difference MP5: differential main peak SP5: differential secondary peak MH1: main Valley SL: auxiliary line MP1, MP2, MP4: main heart rate peak SP1, SP2, SP4: second heart rate peak TP4, TP5: third heart rate peak MH2: main heart rate valley SLS: auxiliary line distance signal t ₁ : first time point t ₂ : the second time point t ₃ : the third time point H1, H2, H4: the valley point H5: the inflection point P ₀ : the amplitude of the first time point P ₃ : the amplitude of the third time point P ₁ : the first control point P ₂ : second control point BC1, BC2: Bezier curve N: unknown point ΔT: the time difference between the main peak of the pulse and the dicrotic wave

Figure 1: A schematic block diagram of a system for implementing the present invention. Fig. 2: a flow chart of the steps of the method for analyzing the arteriosclerosis index of the present invention. Fig. 3A: Waveform diagram of heart rate signal belonging to dicrotic wave. Fig. 3B: Waveform diagram of heart rate signal belonging to dicrotic wave after Fourier transform. Fig. 4A: Waveform diagram of a heart rate signal belonging to a heavy sine wave. Fig. 4B: Waveform diagram of heart rate signal belonging to heavy sine wave after Fourier transform. Figure 5: Waveform diagram of a heart rate signal belonging to a dicrotic wave. Fig. 6: A flow chart of the steps of another method for measuring the arteriosclerosis index without the heavy sine wave of the present invention. FIG. 7A : a waveform diagram of an auxiliary line for establishing a heart rate signal belonging to a dicrotic wave. Fig. 7B: Waveform diagram of the heart rate signal in Fig. 7A after primary differentiation. Fig. 7C: Waveform diagram of the distance signal of the auxiliary line established according to the heart rate signal in Fig. 7A. FIG. 8A : A waveform diagram of an auxiliary line for establishing a heart rate signal belonging to a heavy sine wave. Fig. 8B: Waveform diagram of the heart rate signal in Fig. 8A after primary differentiation. Fig. 8C: Waveform diagram of the auxiliary line distance signal established according to the heart rate signal in Fig. 8A. Fig. 9A: Heart rate signal and auxiliary line waveform diagram of inconspicuous dicrotic wave. Fig. 9B: Waveform diagram of the distance signal of the auxiliary line established according to the heart rate signal in Fig. 9A. FIG. 10A : a heart rate signal belonging to a dicrotic wave and a waveform diagram of a Bezier curve. Fig. 10B: Waveform diagram of the heart rate signal in Fig. 10A after primary differentiation. Fig. 10C: Waveform diagram of the auxiliary line distance signal established according to the heart rate signal in Fig. 10A. Fig. 11A: A heart rate signal belonging to a heavy sine wave and a waveform diagram of a Bezier curve. Fig. 11B: Waveform diagram of the heart rate signal in Fig. 11A after primary differentiation. Fig. 11C: Waveform diagram of the distance signal of the auxiliary line established according to the heart rate signal in Fig. 11A. FIG. 12A : Waveform diagram of the overlapping interval between the heart rate signal belonging to the dicrotic wave and the Bezier curve. FIG. 12B : a waveform diagram of the primary differential in the overlapping interval between the heart rate signal belonging to the dicrotic wave and the Bezier curve. FIG. 13A : a waveform diagram of the overlapping interval between the heart rate signal belonging to the heavy sine wave and the Bezier curve. FIG. 13B : a waveform diagram of the primary differential in the overlapping interval of the heart rate signal belonging to the heavy sine wave and the Bezier curve. FIG. 14A : Heart rate signal and Bezier curve waveform diagram with double beat waves. Fig. 14B: a first differential waveform diagram of the heart rate signal in Fig. 14A. Figure 15: Waveform diagram of a heart rate signal belonging to a dicrotic wave. Fig. 16A: A waveform diagram of a heart rate signal belonging to a heavy sine wave. Figure 16B: The waveform diagram of the heart rate signal in Figure 16A after primary differentiation

10: Wear detection device

11: Control unit

13: Light emitting unit

15: Light receiving unit

20: Action monitoring device

30: Cloud server

Claims (8)

An arteriosclerosis risk assessment system, comprising: a wearable detection device, which is used to detect a heart rate signal of a subject; a motion monitoring device, which is wirelessly connected to the wearable detection device, and the motion detection device is used to receive the heart rate signal, and calculate an arteriosclerosis index, a heart rate variation index, and the ratio of the heavy sine wave according to the heart rate signal, and compare the arteriosclerosis index, the heart rate variation index, and the ratio of the heavy sine wave with an arteriosclerosis index upper limit value, a threshold value of heart rate variability index and an upper limit value of the ratio of heavy sine wave, and then convert it into a risk index according to the comparison result; a cloud server, which is wirelessly connected to the mobile monitoring device, and is used to receive a plurality of the subjects The arteriosclerosis index of each subject is calculated, and the upper limit of the arteriosclerosis index is calculated; wherein the activity monitoring device performs the following steps to obtain the arteriosclerosis index: distinguishing the heart rate signal into continuous complex sub-heart rate signals , and generate a frequency pulse signal through Fourier transform for each sub-heart rate signal; calculate an S value of each sub-heart rate signal, and the S value represents:

; Wherein A _first-f is the intensity value of this frequency pulse signal at double frequency, and A _second-f is the intensity value of this frequency pulse signal at double frequency; if this S value is less than a first threshold value, then compare a time Ratio and a second threshold; wherein, calculate a first time difference and a second time difference of the sub-heart rate signal, and divide the first time difference by the second time difference to obtain the time ratio, the first time difference represents the sub-heart rate The time difference between a heart rate main peak and a heart rate sub-peak of the signal, the second time difference represents the time difference between the heart rate main peak and a main trough of the sub-heart rate signal; calculate the artery according to the height of a subject and the first time difference sclerosis index, wherein the calculation formula of the arteriosclerosis index is:

, where SI is the arteriosclerosis index, h is the height of the subject, and ΔT ₁ is the first time difference. In the arteriosclerosis risk assessment system described in Claim 1, the mobile monitoring device averages multiple arteriosclerosis indices to obtain the average value of the arteriosclerosis indices. The arteriosclerosis risk assessment system as described in claim 2, wherein the motion monitoring device further performs a step of removing heavy sine waves, including: establishing an auxiliary line of the sub-heart rate signal, and the auxiliary line is obtained from the sub-heart rate signal A heart rate main peak linearly extends to a heart rate main valley; wherein the heart rate main peak represents the maximum amplitude of the sub-heart rate signal, and the heart rate main valley represents the lowest amplitude of the sub-heart rate signal; establish an auxiliary line distance signal and calculate the The number of troughs of the auxiliary line distance signal; take a first time point and a second time point of a differential signal, and calculate the distance between the auxiliary line and the sub-heart rate signal between the first time point and the second time point, wherein , the differential signal is a differential waveform of the sub-heart rate signal, the first time point is the time when the first trough of the differential signal along the positive direction of the time axis occurs, and the second time point is the time when the main trough of the heart rate occurs, The auxiliary line distance signal represents the distance function formed by all the vertical distances between the auxiliary line and the differential signal in the interval from the first time point to the second time point; if the number of troughs is 2, it represents the sub-heart rate The signal is a dicrotic wave, and if the number of troughs is 1, it means that the sub-heart rate signal is a dicrosine wave. The arteriosclerosis risk assessment system as described in Claim 3, wherein the action monitoring device further includes the following steps when removing the heavy sine wave: calculating a Bezier curve, the calculation formula of the Bezier curve is: B(t )= P ₀ (1- t ) ³ +3 P ₁ (1- t ) ² t +3 P ₂ (1- t ) t ² + P ₃ t ³ ;

,0

t

1; among them, P ₀ is the first valley point along the positive direction of the time axis after the sub-heart rate signal is differentiated once; P ₃ is the amplitude at a third time point, and the third time point is the time when the main valley of the heart rate occurs ; P ₁ is a first control point, which is obtained by substituting the first trough occurrence time (t ₁ ) of the auxiliary line distance signal along the positive direction of the time axis into the calculation formula of the Bezier curve; P ₂ is a second The control point is obtained by substituting the second trough occurrence time (t ₂ ) of the auxiliary line distance signal along the positive time axis into the calculation formula of the Bezier curve; compare the correlation between the Bezier curve and the heart rate signal , if a correlation coefficient between the Bezier curve and the sub-heart rate signal is less than a third threshold, then calculate the atherosclerosis index of the sub-heart rate signal; wherein, the calculation method of the correlation coefficient is:

; Γ is the correlation coefficient, n is the number of samples, x represents the y-axis coordinate value of the heart rate signal at the same time point; y represents the y-axis coordinate value of the Bezier curve at the same time point. In the arteriosclerosis risk assessment system described in claim 4, when the action monitoring device calculates the Bezier curve, if the auxiliary line distance signal has two troughs, then P ₀ , P ₃ , B(t ₁ ), B(t ₂ ) is substituted into the Bezier curve calculation formula and calculated three times to obtain P ₁ and P ₂ , and then obtain the Bezier curve of the sub-heart rate signal according to P ₁ , P ₂ , P ₀ and P ₃ . In the arteriosclerosis risk assessment system described in claim 4, when the action monitoring device calculates the Bezier curve, if the auxiliary line distance signal has a trough, then P ₀ , P ₃ , B(t ₁ ), B (t ₂ ) into the Bezier curve calculation formula and calculate three times to obtain P ₁ and P ₂ , and then obtain the Bezier curve of the sub-heart rate signal according to P ₁ , P ₂ , P ₀ and P ₃ , The calculation method of B(t ₂ ) is:

In the arteriosclerosis risk assessment system described in claim 5 or 6, when the action monitoring device removes the heavy sine wave, it further includes the following steps: If the auxiliary line distance signal has two troughs, take the first one along the time axis. The amplitude of the heart rate signal is greater than the interval of the amplitude of the Bezier curve, and the sub-heart rate signal in the interval is differentiated once. The first peak of the interval along the positive direction of the time axis is a differential sub-peak, and the differential sub-peak is taken. The arteriosclerosis index is calculated from the time difference between the sub-peak of the heart rate corresponding to the peak occurrence time and the main peak of the heart rate. In the arteriosclerosis risk assessment system described in claim item 7, the action monitoring device further performs the following steps: calculating the ratio of the heavy sine wave; dividing the number of the sub-heart rate signals whose attribute is the heavy sine wave by the total number of the sub-heart rate signals Total, to get the scale of the heavy sine wave.

Images