LU503929B1

LU503929B1 - Single-lead ecg monitoring method and system based on morphological contour algorithm

Info

Publication number: LU503929B1
Application number: LU503929A
Authority: LU
Inventors: Zhaochuan Liu; Lei Tian; Dongying Wang; Pan Tian; Yu Zheng; Heling Huang; Yingkai Cui; Chenyang Wang; Xianghe Wang; Xuebin Cao
Original assignee: Pla Army 82Nd Group Military Hospital
Priority date: 2022-04-14
Filing date: 2023-04-13
Publication date: 2023-10-13
Also published as: CN114795235B; CN114795235A

Abstract

The invention provides a single-lead ECG monitoring method and a system based on morphological contour algorithm, belonging to the technical field of biomedicine. The method includes following steps: collecting initial ECG signals and motion signals in a set time period; based on the adaptive filter, according to the motion signals, removing the motion pseudo signal in the initial ECG signal to obtain a filtered ECG signal; clustering the filtered ECG signals by adopting a morphological contour algorithm of Top-hat transform to obtain a top hat signal; enhancing the top hat signal to obtain an ECG signal sequence; determining a plurality of R peaks according to the ECG signal sequence; according to the time interval between two adjacent R peaks, determining the heart rate of the corresponding time period. Based on adaptive filter and morphological contour algorithm, the motion artifacts and noise signals caused by motion are effectively removed.

Description

DESCRIPTION LU503929

SINGLE-LEAD ECG MONITORING METHOD AND SYSTEM BASED ON

MORPHOLOGICAL CONTOUR ALGORITHM

TECHNICAL FIELD

The invention relates to the technical field of biomedicine, and in particular to a single-lead ECG monitoring method and a system based on morphological contour algorithm.

BACKGROUND

Heart rate is an important parameter to monitor cardiovascular diseases, and its stability directly reflects the quality of heart function, and it is also an important index to guide physical exercise. According to the report of the World Health Organization, the number of people who die of cardiovascular diseases accounts for about one-third of all deaths worldwide every year, ranking first among the population. Monitoring ECG is an important means to diagnose and monitor cardiovascular diseases, but short-term static

ECG detection in hospitals is of little significance to the discovery and prevention of cardiovascular diseases. What is more needed is continuous and long-term dynamic

ECG monitoring in daily life and exercise.

Motion artifact is one of the main noises of ECG (electrocardiogram) signal. When the traditional electrocardiograph is used to measure human ECG signals, people need to lie flat on the bed or even hold their breath, so the interference of motion artifacts is not very serious, which also leads to the fact that the research on eliminating motion artifacts is not as mature as filtering the other three noises of ECG signals (baseline drift, power frequency interference and EMG interference). Due to the appearance of ECG wearable devices in recent years, the dry electrode (knitted electrode) replaces the traditional wet electrode (Ag/AgCl electrode), which has greater contact impedance, becomes more sensitive to human movement and has more obvious motion artifact noise. On the other hand, due to the special needs of wearable devices, wearable ECG, 503929 devices can not restrain the actions of the subjects, and motion artifact noise inevitably exists. In dynamic ECG acquisition, the relative displacement between the electrode and the skin and the stretching of the skin are easy to occur, which leads to the change of electrode-skin impedance, thus leading to greater motion artifact noise. Motion artifact noise is the biggest noise source in wearable ECG monitoring, which makes the potential of electrode skin change as high as several millivolts, thus distorting the signal and seriously interfering with the signal analysis.

Based on the above problems, a novel ECG monitoring method is urgently needed to improve the accuracy of heart rate monitoring in the exercise state.

SUMMARY

The purpose of the invention is to provide a single-lead ECG monitoring method and a system based on morphological contour algorithm, which can improve the accuracy of heart rate monitoring in the exercise state.

In order to achieve the above objectives, the present invention provides the following scheme:

A single-lead ECG monitoring method based on morphological contour algorithm includes: collecting initial ECG signals and motion signals in a set time period; based on the adaptive filter, according to the motion signals, removing the motion pseudo signal in the initial ECG signal to obtain a filtered ECG signal: clustering the filtered ECG signals by adopting a morphological contour algorithm of

Top-hat transform to obtain a top hat signal; enhancing the top hat signal to obtain an ECG signal sequence; determining a plurality of R peaks according to the ECG signal sequence; according to the time interval between two adjacent R peaks, determining the heart rate of the corresponding time period.

Optionally, the initial ECG signal includes initial ECG data at each time; the motion signal includes motion data at each time; the filtered ECG signal includes filtered ECG data at each time; LU503929 the step of based on the adaptive filter, according to the motion signal, removing the motion pseudo signal in the initial ECG signal to obtain a filtered ECG signal, specifically includes:

For time k, according to the motion data at time k and the coefficient of the adaptive filter at time k, determining the coefficient of the adaptive filter at time k+1; presetting the coefficient of the adaptive filter at the initial moment, and k > 0: according to the coefficient of the adaptive filter at time k+1, removing the motion artifacts of the initial ECG data at time k to obtain the filtered ECG data at time k.

Optionally, determining the coefficients of the adaptive filter at time k+1:

Wk +1) = Wk) + LO, ke > 0; where, w(k + 1) is the coefficient of the adaptive filter at time k+1, w(k) is the coefficient of the adaptive filter at time k, u is the step size of the adaptive filter, e(k) is the error of the adaptive filter at time k, x(k) = {rt(k),rt(k — 1) ------,rt(k — M + 1)}, rt(k) is the motion data at time k, M is the order of the filter, k>M>0, ¢ is a constant, and ()" is a transposition operation.

Optionally, determining the filtered ECG data at time Æ: fk) = d(k) —w(k + 1) < rt(k);

Where, f(k) is the filtered ECG data at time k, d(k) is the initial ECG data at time k, W(k + 1) is the coefficient of the adaptive filter at time k+1 and rt(k) is the motion data at time k.

Optionally, determining the top hat signal:

H=f-fOg; where, H is the top hat signal, f is the filtered ECG signal, g is the structural element, f © g reprensents the open operation between fand g.

Optionally, the ECG signal sequence includes enhanced ECG data at each time; determining a plurality of R peaks according to the ECG signal sequence specifically includes: acquiring the maximum value of enhanced ECG data in the ECG signal sequence, and determining a threshold value according to the maximum value;

the initialization flag bit is 0; LU503929 traversing enhanced ECG data in the ECG signal sequence in turn, and setting a flag bit to 1 when the enhanced ECG data is less than a threshold value; when the enhanced ECG data is greater than the threshold and the flag bit is 1, the position of the enhanced ECG data is determined as the position of the R peak, and the flag bit is set to 0.

Optionally, the heart rate is determined by the following formula:

HeartRate = en where, HeartRate is the heart rate, fs is the sampling rate, Rnew — Roua iS the time interval between two adjacent R peaks, R,., and Row are the times corresponding to two adjacent R peaks.

In order to achieve the above purpose, the present invention also provides the following scheme:

A single-lead ECG monitoring system based on morphological contour algorithm comprises an ECG garment, a data acquisition device and a processor; the data acquisition device is arranged at the inner side of the ECG garment, and is used for acquiring the initial ECG signals and motion signals of the human body in real time; the processor is connected with the data acquisition device, and the processor comprises: a pseudo signal removing module connected with the data acquisition device and used for removing the motion pseudo signal in the initial ECG signal based on the adaptive filter and the motion signals to obtain a filtered ECG signal; a clustering module is connected with the pseudo signal removing module and used for clustering the filtered ECG signal by adopting a morphological contour algorithm of

Top-hat transformation to obtain a top-hat signal; an enhancement module is connected with the clustering module and used for enhancing the top hat signal to obtain an ECG signal sequence; a R peak determination module is connected with the enhancement module and used for determining a plurality of R peaks according to the ECG signal sequence; a heart rate determination module is connected with the R peak determination module and used for determining the heart rate of the corresponding time periad 503929 according to the time interval between two adjacent R peaks.

Optionally, the data acquisition device comprises: an electrocardiosignal acquisition component arranged at the inner side of the ECG garment and used for acquiring the electrocardiosignal of a human body in real time; an acceleration circuit is arranged at the inner side of the ECG garment, and is used for collecting the acceleration of a human body in real time to obtain a motion signal.

Optionally, the Electrocardiosignal acquisition component comprises a first metal fabric dry electrode, a second metal fabric dry electrode and a third metal fabric dry electrode: the first metal fabric dry electrode is arranged at the position of chest lead V2; the second metal fabric dry electrode is arranged at the position of 1cm on the left side of the chest lead V2; the third metal fabric dry electrode is arranged at a position 1cm below the right side of the chest lead V2.

According to the specific embodiment provided by the invention, the invention discloses the following technical effects: removing the motion artifacts in the initial ECG signal based on an adaptive filter to obtain a filtered ECG signal, which can eliminate the motion artifacts in a dynamic environment and improve the stability of wearable ECG monitoring equipment in detecting ECG signals in a moving state; then clustering the filtered ECG signal by using a morphological contour algorithm of Top-hat transformation to obtain an ECG signal sequence; and finally, determining a plurality of R peaks according to the ECG signal sequence. Morphological contour algorithm can effectively remove motion artifacts and noise signals caused by motion, and improve the accuracy of heart rate monitoring.

BRIEF DESCRIPTION OF THE FIGURES LU503929

In order to explain the embodiments of the present invention or the technical scheme in the prior art more clearly, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For ordinary people in the field, other drawings can be obtained according to these drawings without paying creative labor.

FIG. 1 is a flow chart of a single-lead ECG monitoring method based on morphological contour algorithm of the present invention;

FIG. 2 is a schematic diagram of the module structure of a single-lead ECG monitoring system based on morphological contour algorithm of the present invention;

FIG. 3 is a schematic structural diagram of a single-lead ECG monitoring system;

FIG. 4 is a schematic diagram of the connection relationship of the magnetic interface circuit;

FIG. 5 is a flow chart of data acquisition using ADS1292R;

FIG. 6 is a flow chart of the use of the single-lead ECG monitoring system;

FIG. 7(a) is a schematic diagram of the initial ECG signal collected by the single-lead ECG monitoring system when the heart rate is 72 beats/min;

FIG. 7(b) is a schematic diagram of the signal processed by the morphological contour algorithm in FIG. 7(a);

FIG. 7(c) is a schematic diagram of the initial ECG signal collected by the single-lead ECG monitoring system when the heart rate is 118 beats/min;

FIG. 7(d) is a schematic diagram of the signal processed by the morphological contour algorithm in FIG. 7(c).

Symbol descriptions:

ECG garment-1, data acquisition device-2, electrocardiosignal acquisition component -21, irst metal fabric dry electrode-211, second metal fabric dry electrode-212, third metal fabric dry electrode-213, acceleration circuit-22, processor-3, pseudo signal removing module-31, clustering module-32, enhancement module-33, R peak determination module-34, heart rate determination module-35, power supply circuit-4,

Bluetooth module-5, housing-6, magnetic interface circuit-7.

DESCRIPTION OF THE INVENTION LU503929

In the following, the technical scheme in the embodiment of the invention will be clearly and completely described with reference to the attached drawings. Obviously, the described embodiment is only a part of the embodiment of the invention, but not the whole embodiment. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in the field without creative labor belong to the scope of protection of the present invention.

The purpose of the invention is to provide a single-lead ECG monitoring method and a system based on morphological contour algorithm. Through the adaptive filter and morphological contour algorithm, motion artifacts and noise signals brought by motion are effectively removed, and the accuracy and stability of heart rate monitoring are improved.

In order to make the above objects, features and advantages of the present invention more obvious and easier to understand, the present invention will be further described in detail with the attached drawings and specific embodiments.

As shown in FIG. 1, the single-lead ECG monitoring method based on the morphological contour algorithm of the present invention includes:

S1: collecting initial ECG signals and motion signals in a set time period. In this embodiment, the initial ECG signals and motion signals are collected after the human body wears the ECG garment under the motion state of 5 km/h-10 km/h.

S2, based on the adaptive filter, according to the motion signals, removing the motion pseudo signal in the initial ECG signal to obtain a filtered ECG signal.

S3, clustering the filtered ECG signal by adopting the morphological contour algorithm of Top-hat transform to obtain a top hat signal. Specifically, the top hat signal includes clustered ECG signals at each time.

S4, enhancing the top hat signal to obtain an ECG signal sequence. The ECG signal sequence includes enhanced ECG data at each time.

Specifically, the top hat signal can improve the signal-to-noise ratio through an enhancement operation, which can amplify the R wave extracted after the Top-hat transform and reduce the unnecessary noise, and obtain the morphological category of

QRS wave and the ECG signal sequence S: LU503929

S = 1000 x (H(k + ky) — H(k)) - (H(k + ky) — H(k + ky + k2));

Where, H(k) is the clustered ECG signal at time k in the top hat signal, k, = (N, + 1)/2 k, = N,+1- k,, N, represent the length of structural element g. The position of

R peak in ECG signal sequence S is more obvious, and then the negative value and the value close to 0 can be directly cleared, so as to further improve the signal-to-noise ratio.

S5, determining a plurality of R peaks according to the ECG signal sequence.

S6, determining the heart rate of the corresponding time period according to the time interval between two adjacent R peaks. Specifically, the following formula is adopted to determine the heart rate:

HeartRate = en where, HeartRate is the heart rate, fs is the sampling rate, Rnew — Roua iS the time interval between two adjacent R peaks, R,., and Row are the times corresponding to two adjacent R peaks. The sampling rate fs is the hardware sampling rate, and it is also the reciprocal of the time interval between two adjacent acquisitions, indicating how many data are acquired per second. In this embodiment, the sampling rate is 540Hz. the analog (continuous) ECG signal is digitized (discretized).

Further, the initial ECG signal includes initial ECG data at each time; the motion signal includes motion data at each time; the filtered ECG signal includes filtered ECG data at each time; S2 specifically includes:

S21: For time k, according to the motion data at time k and the coefficient of the adaptive filter at time k, determining the coefficient of the adaptive filter at time k+1; presetting the coefficient of the adaptive filter at the initial moment. Specifically, the following formula is adopted to determine the coefficient of the adaptive filter at time k+1:

Wk +1) = Wk) + LO, ke > 0; where, w(k + 1) is the coefficient of the adaptive filter at time k+1, w(k) is the coefficient of the adaptive filter at time k, u is the step size of the adaptive filter, which is related to the stability and convergence rate of LMS algorithm, e(k) is the error of the adaptive filter at time k, X(k) = {rt(k),rt(k — 1) ------,rt(k — M + 1)}, x(k) is the input reference sequence of adaptive filter, ()" is a transposition operation, rt(k) is the motion data at time k, M is the order of the filter, that is, the total number of motion data IN )503929 motion signal, k>M>0, ¢ is a constant, which avoids the divergence of the algorithm caused by the denominator being zero during the algorithm operation.

Specifically, constantly updating the coefficients of the adaptive filter according to the motion signal, and the coefficients with the smallest mean square error E{e?(k)} are selected according to the cost function by using the criterion of Wiener optimal filter.

S22: According to the coefficient of the adaptive filter at time k+1, the motion artifacts of the initial ECG data at time k are removed to obtain filtered ECG data at time k. Specifically, the filtered ECG data at time k is determined by the following formula: fk) = d(k) — W(k +1) rt (k);

Further, in S3, the following formula is adopted to determine the top hat signal:

H=f-fOg; where, H is the top hat signal, f is the filtered ECG signal, g is the structural element, f © g represents the open operation between f and g.

Specifically, the open operation is corrosion first and then expansion, and the process is as follows: fOg=(09Dg=hrDy; f © g means that g expands f, h@ g means that g expands h, h=fOg.

When g is a horizontal organization element, the open operation will eliminate the peak in the filtered ECG signal Fand get the base signal of the filtered ECG signal f. f@9= dpi, We +m) — g(m)} (k = 0,1,..., Ny — Np). h®g= RUC + m) + g(m)} (k = 0,1,..., N, — Ng). where, N, represents the filtered ECG signal length of f, and f(k+m) represents the filtered ECG data at time k+m, g(m) represents the m-th data in the structural elements, and N, represents the length of the structural element g, actually, N, > Nj.

Because the structural element g has a length, the effective signal length will reduce the corresponding length N, of the structural element after morphological operation.

The result of corrosion operation always takes the minimum value of filtered ECG signal 503929 in the corresponding interval of structural elements.

Taking g(m) = 0, the result of the open operation is only related to its length, simplifying the operation. Taking N, as:

Ng = [E28] + 1;

Where [ ] stands for rounding. According to the width of R wave, the value of Tr generally varies from 25 - 40 ms.. f; is the sampling rate in Hz. N, indicates the length of the structural element g, which is a template used for morphological processing of sequences. Structural elements with different lengths have different processing effects.

The larger the N, is, the wider the width of the top hat is. In order to ensure the top hat with R wave length, it is necessary to adjust the length of structural elements through the time width of R wave.

Further, S4 specifically includes:

S41: Obtaining the maximum value Smax of enhanced ECG data in the ECG signal sequence, and determining the threshold value according to the maximum value

Sthresnoia- Specifically,

S42: The initialization flag bit is O.

S43: sequentially traversing the enhanced ECG data in the ECG signal sequence, and setting a flag bit to 1 when the enhanced ECG data is less than a threshold; when the enhanced ECG data is greater than the threshold and the flag bit is 1, the position of the enhanced ECG data is determined as the position of the R peak, and the flag bit is set to 0.

Specifically, traversing the ECG signal sequence in time, constantly comparing it with the threshold S*hreshoia, ANd finding the position of R wave peak again when the enhanced ECG data in the ECG signal sequence is less than the threshold Senreshotd) that is findRFlag=1. When the enhanced ECG data is greater than the threshold

Sthreshoia @Nd findRFlag=1, recording the time corresponding to the enhanced ECG data, which is a position of the R peak, and making findRFlag=0 at the same time. This cycle is repeated until the end of traversing the ECG signal sequence. In the cycle, the found R peak positions are stored in a sequence, and two adjacent R peak positions are a pair of old and new R peaks.

As shown in FIG. 2, the single-lead ECG monitoring system based on morphological 503929 contour algorithm of the present invention includes an ECG garment 1, a data acquisition device 2 and a processor 3.

Specifically, the main body of ECG garment 1 is made of flexible, lightweight, high elasticity and good air permeability natural fiber fabric, and is made of knitted structure.

The data acquisition device 2 is arranged inside the ECG garment 1, and is used for acquiring the initial ECG signal and motion signal of the human body in real time.

The processor 3 is connected with the data acquisition device 2, and the processor 3 comprises a pseudo signal removing module 31, a clustering module 32, an enhancement module 33, an R peak determination module 34 and a heart rate determination module 35.

Wherein, the pseudo signal removing module 31 is connected with the data acquisition device 2, and the pseudo signal removing module 31 is used for removing the motion false signal in the initial ECG signal based on the adaptive filter and according to the motion signal to obtain a filtered ECG signal.

The clustering module 32 is connected with the pseudo signal removing module 31, and the clustering module 32 is used for clustering the filtered ECG signal by using the morphological contour algorithm of Top-hat transform to obtain a top-hat signal.

The enhancement module 33 is connected with the clustering module 32, and the enhancement module 33 is used for enhancing the top hat signal to obtain an ECG signal sequence.

The R peak determination module 34 is connected with the enhancement module 33, and the R peak determination module 34 is used for determining a plurality of R peaks according to the ECG signal sequence.

The heart rate determination module 35 is connected with the R peak determination module 34, and the heart rate determination module 35 is used to determine the heart rate of the corresponding time period according to the time interval between two adjacent R peaks.

In this embodiment, the processor 3 adopts an STM32 single chip microcomputer.

Specifically, the initial ECG signal includes initial ECG data at each time; the motion signal includes motion data at each time; the filtered ECG signal includes filtered ECG data at each time. LU503929

The pseudo signal removing module 31 includes a coefficient updating unit and a removing unit. The coefficient updating unit is connected with the data acquisition device 2, and is used for determining the coefficient of the adaptive filter at time k+1 according to the motion data at time k and the coefficient of the adaptive filter at time k; the coefficient of the adaptive filter at the initial time is preset, and k>0.

The removing unit is connected with the coefficient updating unit, and the removing unit is used for removing the motion pseudo signal of the initial ECG data at time k according to the coefficient of the adaptive filter at time k+1 time to obtain the filtered

ECG data at time K.

Further, the ECG signal sequence includes enhanced ECG data at each time. The

R peak determination module 34 includes a threshold determination unit, a flag bit initialization unit and a peak position determination unit.

Wherein, the threshold determining unit is connected with the enhancement module 33, and the threshold determining unit is used for acquiring the maximum value of enhanced ECG data in the ECG signal sequence, and determining the threshold value according to the maximum value.

The flag bit initialization unit initializes the flag bit to O.

The peak position determining unit is respectively connected with the threshold value determining unit and the flag bit initializing unit, and the peak position determining unit is used for sequentially traversing the enhanced ECG data in the ECG signal sequence and setting the flag bit to 1 when the enhanced ECG data is less than the threshold value; when the enhanced ECG data is greater than the threshold and the flag bit is 1, the position of the enhanced ECG data is determined as the position of the R peak, and the flag bit is set to 0.

Further, the data acquisition device 2 includes an ECG signal acquisition part 21 and an acceleration circuit 22.

Wherein, the ECG signal acquisition part 21 is arranged inside the ECG garment 1, and the ECG signal acquisition part 21 is used for acquiring the ECG signal of the human body in real time.

Specifically, as shown in FIG. 3, the Electrocardiosignal acquisition component 2503929 includes a first metal fabric dry electrode 211, a second metal fabric dry electrode 212 and a third metal fabric dry electrode 213. The first metal fabric dry electrode 211 is arranged at the position of chest lead V2; the second metal fabric dry electrode 212 is arranged at the position of 1cm on the left side of the chest lead V2; the third metal fabric dry electrode 213 is arranged at a position 1cm below the right side of the chest lead V2.

In this embodiment, three metal fabric dry electrodes are made of high-conductivity silver wires according to the island-bridge structure pattern, and their shapes are circular with a diameter of 2.5cm.

The acceleration circuit 22 is arranged at the inner side of the ECG garment 1, and is used for collecting the acceleration of the human body in real time to obtain a motion signal. In this embodiment, the acceleration circuit 22 adopts a capacitive ADXL345 acceleration sensor. ADXL345 acceleration sensor can measure both dynamic acceleration caused by movement or impact and static acceleration, and can be used as a tilt sensor to judge human posture. At the same time, an analog-to-digital converter is integrated to convert the triaxial acceleration analog signal into a digital signal, and the data is transmitted with the processor 3 through the 12C interface.

Further, the processor 3 also includes a data storage module. The data storage module is respectively connected with the data acquisition device 2 and the heart rate determination module 35, and is used for storing initial ECG signals, motion signals and heart rates. In this embodiment, the data storage module adopts a MicroSD flash memory card with a memory of 128GB and a size of 15mmx11mmx1mm, and the communication mode of the data storage module adopts SDIO interface communication.

Furthermore, the single-lead ECG monitoring system also includes a power supply circuit 4. The power supply circuit 4 is respectively connected with the data acquisition device 2 and the processor 3. In this embodiment, the power supply circuit 4 uses 125mA and 3.7V rechargeable lithium batteries to meet the requirements of long-term wearing and continuous monitoring of equipment. Specifically, TLV700 linear voltage regulator chip is used for voltage conversion, and the voltage of power supply circuit 4 is converted into 3.3V and +2.5V. 3.3V voltage supplies power for STM32 main control module and data storage module. The analog power supply voltage of ADS1292R is

-2.5V, and the data power supply voltage is 3.3V. LU503929

In addition, the single-lead ECG monitoring system also includes a Bluetooth module 5. In this embodiment, the Bluetooth module 5 adopts HC-42 Bluetooth circuit.

In order to protect the hardware circuit, the single-lead ECG monitoring system of the present invention also includes a housing 6. The processor 3 and the Bluetooth module 5 are both arranged inside the housing 6. In this embodiment, the overall dimension of the housing 6 is 3.2cmx4cmx=0.8cm.

As a specific embodiment, as shown in FIG. 4, the single-lead ECG monitoring system further includes a magnetic interface circuit 7. The processor 3, the power supply circuit 4, and the Bluetooth module 5 are all arranged on the electrocardiograph 1 through a magnetic interface circuit 7.

As another embodiment, the ECG signal acquisition part 21 adopts the ADS1292R

ECG acquisition front end. In this embodiment, the ADS1292R ECG acquisition front-end adopts the ADS1292R chip of TI company. IN1P/IN1N of ADS1292R chip are input ports, and differential input is adopted to reduce common-mode interference.

CLKSEL is a clock pin, which can select an external clock and an internal clock. Because the internal clock uses an internal oscillator circuit on the chip, it is greatly affected by temperature and its accuracy is not high. Therefore, the invention selects an external clock as a clock signal, connects the CLKSEL to ground, and inputs a 2MHz clock signal to the CLKSEL pin. ADS1292R is built with two Pmgrammable Gain Amplifier (PGA), which use channel 1 to collect ECG signals, and set PGA1=2, ADC sampling rate to 500SPS, and internal reference voltage VREF to 2.42V. At the same time, ADS1292R can suppress common-mode interference of human body through "right leg drive circuit".

As shown in FIG. 5, the process of data acquisition using ADS1292R ECG acquisition front-end is shown.

The invention adopts vest-type ECG clothes, realizes the monitoring of heart rate in the exercise state, can resist long-term sweat corrosion and has good air permeability and skin friendliness. ECG clothes are light, soft, sensitive, comfortable to wear and simple to use. The single-lead ECG monitoring system of the invention is small in size, thin in thickness, light in weight, low in power consumption, comfortable to wear and high in accuracy, and can be used for resisting motion interference under human wear.

In order to better understand the scheme of the present invention, the flow of the 503929 single-lead ECG monitoring system will be further explained with specific examples.

As shown in FIG. 6, firstly, the system is initialized, and ECG signals and motion signals are collected. After the collection, the ECG signals are filtered, baseline removed and stored. When the acquisition times are more than 150 times, it is judged whether to transmit through Bluetooth, if so, the ECG signal is transmitted to the upper computer, otherwise, the ECG signal is written into the memory card, and the ECG signal is continuously collected.

In order to better understand the scheme of the present invention, the use flow of the single-lead ECG monitoring system based on morphological contour algorithm of the present invention will be further explained with specific examples.

Step 1: Determining the material, shape and size of metal fabric dry electrode.

Step 2: Determining the material and size of the ECG garment and the specific position of the metal fabric dry electrode in the ECG garment.

Step 3: Hardware design of the single-lead ECG monitoring system, which mainly includes ADS1292 acquisition module, 8G Micro-SD data storage circuit, 125mA power supply circuit, HC-42 Bluetooth circuit, triaxial acceleration circuit and magnetic interface circuit.

Step 4: Wearing single-lead ECG monitoring equipment and ECG monitoring electrodes of Mindray PM-9000 monitor, and carrying out squat and 5km/h-10km/h running. After the power-on reset of the single-lead ECG monitoring system, the system clock configuration, interrupt priority grouping configuration, and initialization of all peripherals are carried out. After the data acquisition of ADS1292R is completed, an interrupt is triggered. The processor receives data in response to the interrupt, reads the collected data through direct memory access DMA, performs conversion and denoising, writes the processed data into the memory card, and then transmits it to HC-42

Bluetooth module through serial port. After the hardware is powered on, the hardware initialization is completed, and the sampling frequency of ADC (analog to digital converter), the amplification factor of PGA (Programmable Gain Amplifier), reference voltage selection, clock selection and other operations are set through the configuration register group.

Step 5: Single-lead ECG monitoring equipment collects ECG signals and motion 503929 signals in real time, and the adaptive filter adjusts the parameters of the filter in real time according to the collected ECG signals and motion signals, so as to remove motion artifacts and noise interference signals and output the processed signals.

Step 5: performing top-hat transform on the output signal in step 4 to obtain the

ECG signal sequence.

Step 6: calculating the real-time heart rate according to the ECG signal sequence.

Step 7: analyzing the error of the heart rate calculated in step 6 and the heart rate measured by wearing Mindray PM-9000 monitor during exercise, and calculating the accuracy of heart rate measured by single-lead ECG monitoring equipment.

In addition, the invention tests the single-lead ECG monitoring system in the following two ways to verify the accuracy of heart rate measurement.

Test 1: The standard ECG signal source is SKX-2000 ECG signal analog generator, which generates 72 beats/Min and 118 abnormal ECG signals (arrhythmia) respectively.

The test results show that the single-lead ECG monitoring system provided by the invention can accurately collect normal and abnormal ECG signals generated by the

ECG signal generator, and accurately calculate the heart rate, with an accuracy rate of 100%. The results are shown in Figures 7(a), 7(b), 7(c) and 7(d).

Test 2:

The wearer wears the single-lead ECG monitoring system of the present invention and Mindray PM-9000 monitor at the same time, and measures the ECG waveform and heart rate of the wearer after different exercise times in a calm state, after squatting for 1 minute and at a speed of 5km/h respectively. The heart rate measured by the single-lead

ECG device of the present invention is compared with the heart rate measured by

Mindray PM-9000 monitor, and the error analysis shows that the average error of heart rate measurement is 7.35%, and the heart rates of two different testers in different states are shown in Table 1.

Table 1 Heart rate measured by single-lead ECG monitoring system and Mindray 503929

PM-9000 monitor (beats/minute)

Different States PM-9000 Single-lead ECG | Relative error% monitoring system 1min after 123 132 7.32 exercise 2min after 111 120 8.11 exercise 3min minutes 105 110 4.76 after exercise 4min after 94 102 8.51 exercise 5min minutes 88 95 7.95 after exercise

Gmin after 91 7.69 exercise

CI em 1min after 116 110 5.17 exercise 2min after 94 103 9.57 exercise 3min minutes 94 9.30 after exercise 4min after 88 10.00 exercise

ERIC em

6min after | 79 85 7.59 exercise

Each embodiment in this specification is described in a progressive way, and each embodiment focuses on the differences from other embodiments, so it is only necessary to refer to the same and similar parts between each embodiment. As for the system disclosed in the embodiment, because it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points can only be described in the method part.

In this paper, specific examples are used to explain the principle and implementation of the invention, and the description of the above examples is only used to help understand the method and its core idea of the invention; At the same time, for ordinary technicians in this field, according to the idea of the invention, there will be changes in the specific implementation and application scope. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

CLAIMS LU503929

1. À single-lead ECG monitoring method based on morphological contour algorithm, characterized in that the single-lead ECG monitoring method based on morphological contour algorithm comprises: collecting initial ECG signals and motion signals in a set time period, based on the adaptive filter, according to the motion signals, removing the motion pseudo signal in the initial ECG signal to obtain a filtered ECG signal; clustering the filtered ECG signals by adopting a morphological contour algorithm of Top-hat transform to obtain a top hat signal; enhancing the top hat signal to obtain an ECG signal sequence: S=1000 x (H(k+k1)-H(K))- (H(k+k1)-H(k+k1+k2)); where S is the ECG signal sequence, H(k) is the cluster ECG signal at time k in the top hat signal, ki = (Ng+1)/2, ko = Ng+1-k4, and Ng represents the length of the structural element; determining a plurality of R peaks according to the ECG signal sequence; according to the time interval between two adjacent R peaks, determining the heart rate of the corresponding time period.

2. The single-lead ECG monitoring method based on morphological contour algorithm according to claim 1, characterized in that the initial ECG signal includes initial ECG data at each time; the motion signal includes motion data at each time; the filtered ECG signal includes filtered ECG data at each time; based on the adaptive filter, according to the motion signal, removing the motion pseudo signal in the initial ECG signal to obtain a filtered ECG signal, specifically includes: for time k, according to the motion data at time k and the coefficient of the adaptive filter at time k, determining the coefficient of the adaptive filter at time k+1; presetting the coefficient of the adaptive filter at the initial moment, and k > 0: according to the coefficient of the adaptive filter at time k+1, removing the motion artifacts of the initial ECG data at time k to obtain the filtered ECG data at time k.

3. The single-lead ECG monitoring method based on morphological contour 503929 algorithm according to claim 2, characterized in that the following formula is adopted to determine the coefficient of the adaptive filter at time k+1: Wk + 1) = Wk) + LO, ke > 0; where, w(k + 1) is the coefficient of the adaptive filter at time k+1, w(k) is the coefficient of the adaptive filter at time k, u is the step size of the adaptive filter, e(k) is the error of the adaptive filter at time k, x(k) = {rt(k),rt(k — 1) ------,rt(k — M + 1)}, rt(k) is the motion data at time k, ()" is a transposition operation, M is the order of the filter, k>M>0, € is a constant.

4. The single-lead ECG monitoring method based on morphological contour algorithm according to claim 2, characterized in that the filtered ECG data at time k is determined by the following formula: fk) = d(k) — W(k +1) rt (k); where, f(k) is the filtered ECG data at time k, d(k) is the initial ECG data at time k, W(k + 1) is the coefficient of the adaptive filter at time k+1 and rt(k) is the motion data at time k.

5. The single-lead ECG monitoring method based on morphological contour algorithm according to claim 1, characterized in that the top hat signal is determined by the following formula: H=f-fOg; where, H is the top hat signal, f is the filtered ECG signal, g is the structural element, f © g reprensents the open operation between fand g.

6. The single-lead ECG monitoring method based on morphological contour 503929 algorithm according to claim 1, characterized in that the ECG signal sequence includes enhanced ECG data at each time: determining a plurality of R peaks according to the ECG signal sequence specifically includes: acquiring the maximum value of enhanced ECG data in the ECG signal sequence, and determining a threshold value according to the maximum value; the initialization flag bit is 0; traversing enhanced ECG data in the ECG signal sequence in turn, and setting a flag bit to 1 when the enhanced ECG data is less than a threshold value; when the enhanced ECG data is greater than the threshold and the flag bit is 1, the position of the enhanced ECG data is determined as the position of the R peak, and the flag bit is set to

0.

7. The single-lead ECG monitoring method based on morphological contour algorithm according to claim 1, characterized in that the following formula is adopted to determine the heart rate: HeartRate = en where, HeartRate is the heart rate, fs is the sampling rate, Rnew — Roua iS the time interval between two adjacent R peaks, R,., and Row are the times corresponding to two adjacent R peaks.

8. A single-lead ECG monitoring system based on morphological contour algorithm, 503929 characterized in that the single-lead ECG monitoring system based on morphological contour algorithm comprises an ECG garment, a data acquisition device and a processor;

the data acquisition device is arranged at the inner side of the ECG garment, and is used for acquiring the initial ECG signals and motion signals of the human body in real time;

the processor is connected with the data acquisition device, and the processor comprises:

a pseudo signal removing module connected with the data acquisition device and used for removing the motion pseudo signal in the initial ECG signal based on the adaptive filter and the motion signals to obtain a filtered ECG signal;

a clustering module is connected with the pseudo signal removing module and used for clustering the filtered ECG signal by adopting a morphological contour algorithm of Top-hat transformation to obtain a top-hat signal; the top hat signal includes clustered ECG signals at each time;

an enhancement module is connected with the clustering module and used for enhancing the top hat signal to obtain an ECG signal sequence: S = 1000%(H(k+k1)-H(K)):(H(k+k1)-H(k+k1+k2)); where S is the ECG signal sequence, H(k) is the cluster ECG signal at time k in the top hat signal, k1=(Ng+1)/2, k=Ng+1-k1, and Ng represents the length of the structural element;

a R peak determination module is connected with the enhancement module and used for determining a plurality of R peaks according to the ECG signal sequence;

a heart rate determination module is connected with the R peak determination module and used for determining the heart rate of the corresponding time period according to the time interval between two adjacent R peaks.

9. The single-lead ECG monitoring system based on morphological contour 503929 algorithm according to claim 8, characterized in that the data acquisition device comprises: an electrocardiosignal acquisition component arranged at the inner side of the ECG garment and used for acquiring the electrocardiosignal of a human body in real time; an acceleration circuit is arranged at the inner side of the ECG garment, and is used for collecting the acceleration of a human body in real time to obtain a motion signal.

10. The single-lead ECG monitoring system based on morphological contour algorithm according to claim 9, characterized in that the electrocardiosignal acquisition component comprises a first metal fabric dry electrode, a second metal fabric dry electrode and a third metal fabric dry electrode; the first metal fabric dry electrode is arranged at the position of chest lead V2; the second metal fabric dry electrode is arranged at the position of 1cm on the left side of the chest lead V2; the third metal fabric dry electrode is arranged at a position 1cm below the right side of the chest lead V2.