CN205268157U - Non -contact heart electric sensor and wearable multichannel electrocardio sampling underwear thereof - Google Patents

Abstract

The non-contact ECG sensor and its wearable multi-channel ECG sampling underwear relate to the measurement of bioelectrical signals of the human body or various parts of the human body, in particular to a sensor for electrocardiography and a wearable multi-channel electrocardiogram using the sensor. Channel ECG sampling underwear, the ECG sensor includes the detection element formed inside the PCB board and the ECG signal amplification module mounted on the PCB board; the bottom wiring layer of the PCB board is divided into the detection electrode and the shielding ring surrounding the detection electrode; Via holes evenly distributed along the shielding ring connect the shielding ring and the shielding layer to form a three-dimensional shielding cavity; the detection electrodes obtain ECG signals through capacitive coupling in a non-contact manner. A number of individual small-sized non-contact ECG sensors are distributed inside the ECG sampling underwear, and are connected to the host of the ECG instrument in a multi-channel differential input mode. In the daily electromagnetic environment with wireless signal interference, the collected ECG signals It has the signal quality comparable to the traditional measurement method of professional medical electrocardiograph.

Description

Non-contact ECG sensor and its wearable multi-channel ECG sampling underwear

technical field

The utility model relates to the measurement of bioelectric signals of a human body or each part of a human body, in particular to a sensor for electrocardiography and wearable multi-channel electrocardiographic sampling underwear using the sensor.

Background technique

Heart disease is a slow process, and it is difficult for normal people to perceive small changes in the heart. And it is often sudden, transient and very dangerous, so patients must be able to monitor ECG in real time and conveniently. For some high-risk patients, real-time monitoring and early warning are required to avoid danger. However, existing consumer ECG products are still unable to obtain ECG signals similar to professional dynamic multi-channel electrocardiographs (such as Holter), and ECG signals similar to Holter are the medical standards for diagnosing heart diseases and cardiac physiological conditions. At present, some consumer products related to ECG have appeared in the markets of North America, Europe, and mainland China, such as smart watches that output real-time heart rate, that is, the number of heartbeats per minute, and bra-type or chest-band heart rate monitors, which can be estimated by blood pulse changes. A single-finger optical heart rate monitor for heart rate, a two-finger electrocardiometer that uses one finger on the left and right to obtain ECG signals, and a single-channel electrocardiometer that uses a traditional wet electrode attached to the chest to obtain ECG signals, etc. Compared with the high-cost professional medical electrocardiograph (Holter) at the level of RMB 10,000, these consumer ECG products at the level of hundreds to thousands of yuan are undoubtedly more suitable for the general public, and they are easy to wear and can output signals in real time Its characteristics are superior to those of Holter, so it can greatly improve the public's awareness of their own heart health. Among these ECG-related consumer products, only the latter two can output real-time ECG signals, so they are the closest to Holter. However, since the two-finger ECG requires the user to use both hands to complete the recording, which affects the user's daily life and work, only discontinuous ECG signals can be obtained, and the output signal is limited by the fact that the electrodes are located at the finger and away from the heart. Limited quality. Moreover, the single-channel electrocardiograph has limited anti-interference ability, for example, it cannot effectively capture the heartbeat and monitor clear ECG signals when the user is exercising. It can be seen that since these consumer products cannot obtain high-quality multi-channel ECG signals, the information they obtain has limited medical and physiological significance, and is not enough for diagnosing heart diseases and monitoring cardiac physiological conditions.

Chinese invention patent "a new type of intelligent ECG test health device" (invention patent number: ZL201010539911.3 authorized announcement number: CN102462494B) discloses a new type of intelligent ECG test health device, which obtains ECG signals through biological contact electrodes, adopts The signal processing chip performs amplification, A/D conversion, drift suppression, AC filtering, ECG filtering and contact detection on the ECG signal to obtain high-quality ECG spectrum. Through the communication interface module, it is connected to the computer software analysis and processing device installed on the personal computer, notebook computer, netbook, and mobile phone to record and analyze the ECG waveform for reference in clinical examination. , identification, mood test and other functions. The technical solution adopts a double variable threshold method (namely, a peak threshold and a valley threshold) for judgment and processing. However, in the case of strong interference, such as an unstable baseline of the collected signal level, obvious EMG signals, or electronic device noise caused by motion, the threshold method obviously cannot guarantee the efficiency of capturing ECG signals. Furthermore, the thresholding method often does not capture the moment precisely. Because under noise interference, the threshold value cannot clearly reflect the position relative to the R wave, that is, sometimes it is far away from the R wave, and sometimes it is close, so there is a big error in the heart rhythm estimation. Even if the threshold can be reset, it cannot meet the complex and demanding actual situation.

Chinese invention patent "A mobile terminal and ECG monitoring method for non-contact ECG monitoring" (invention patent number: ZL201110328284.3 authorized announcement number: CN102512153B) discloses a non-contact ECG monitoring mobile terminal and ECG monitoring method An electrical monitoring method, the ECG monitoring method includes: a mobile terminal collects a signal containing ECG data in a non-contact manner; the mobile terminal performs denoising and separate processing of the ECG data in the signal from other data; The mobile terminal extracts ECG data from the separated and processed data; the mobile terminal sends the extracted ECG data to a designated receiving device through a wireless mobile communication network. This prior art solution adopts the method of wavelet transformation, first filters out the noise, and then reconstructs the electrocardiographic signal. Thereby subtly reducing the impact of noise on capturing heartbeat. However, since filtering and reconstruction will have a certain impact on the original signal, this may cause distortion of the ECG signal, so that high-quality ECG signals cannot be obtained, and it is difficult to be used for diagnosing heart diseases and monitoring cardiac physiological conditions.

The Chinese invention patent "a non-contact ECG sensor and its application" (invention patent number: ZL201210128390.1 authorized announcement number: CN102657524B) discloses a non-contact ECG sensor, which uses circular double-sided PCB board electrodes, There are three areas on one surface, and the three areas are electrically insulated. The central circular area is the induction sheet, and the annular shielding area and the annular ground line area are arranged outwards in turn. There are also three areas on the other surface. The central circle is Copper-clad area, the center of the copper-clad area is equipped with a chip welding area including a pre-operational amplifier and a front-end filter, and the periphery of the copper-clad area is a ring-shaped ground area. The copper-clad area on one surface is symmetrical, and the ring-shaped ground wire areas on the two surfaces are symmetrical. The ring-shaped ground area on this surface is connected, and the input of the pre-operational amplifier is connected to the output of the sensor chip. The electrodes of the double-sided PCB board are positioned on the open end of the circular metal shielding box, the side with the induction sheet faces the opening, and the side with the copper-clad area faces the inner cavity of the shielding box. The induction sheet of this technical solution is exposed, and the user needs to wear an underwear or a T-shirt to achieve relative "insulation". Since the insulation of the clothes will be greatly reduced (to the order of K ohms or even lower) under certain conditions (such as user sweating or air humidity), the method of this technical solution is not strictly electrical insulation. In the ECG measurement, the contact between the skin and the electrode will not only cause the polarization voltage, but also the contact between the electrode and the skin will have a polarization resistance. Seen as the source resistance of the entire circuit system, and the input resistance of the preamplifier circuit for voltage division, the changing polarization resistance will also cause the voltage divider output of the preamplifier circuit to be in an unstable state.

Chinese invention patent application "Remote ECG Monitoring and Diagnosis System" (invention patent application number: 201510096714.1 publication number: CN104644159A) discloses a monitoring and diagnosis system, including an ECG data acquisition unit for collecting and storing ECG signals, The ECG data acquisition unit is connected to a mobile terminal for receiving and displaying ECG signals, and the mobile terminal is connected to an ECG analysis platform through a network; the ECG analysis platform can analyze the ECG signals transmitted by the mobile terminal deal with. In order to miniaturize the system, the above-mentioned existing technical solutions all adopt no local storage or small-capacity local storage, or use the data processing method of low-precision short-term recording (reducing data storage pressure), or collect the ECG data through wireless The mode is transmitted to the external receiving device for storage. However, if the user's location is beyond the coverage area of the wireless signal, or if the on-site interference is serious, the above-mentioned prior art solution is likely to cause permanent data loss, and it is difficult to ensure the integrity of the real-time data.

On the other hand, the traditional wet electrode adopts a passive design, and its volume and installation area are small (about 1 cm in diameter), while the non-contact electrode usually adopts an active design, which has a large volume and installation area, which increases the electrode placement. difficult, reducing user comfort. In addition, the input impedance of non-contact electrodes is G ohm level, which is very sensitive to environmental noise and can easily cause saturation of the back-end signal conditioning circuit. Therefore, how to reduce the volume and installation area of non-contact electrodes and improve the shielding effect of electrodes , Improving the signal-to-noise ratio of the input ECG signal and improving the signal quality are also technical problems that must be solved for wearable real-time multi-channel non-contact ECG instruments.

Utility model content

The purpose of the utility model is to provide a wearable real-time multi-channel non-contact electrocardiogram sensor, which can be obtained at low cost, has information medical physiological significance, and can be used for diagnosing heart diseases and monitoring cardiac physiological conditions. High-quality multi-channel ECG signals.

The technical solution adopted by the utility model to solve the problems of the technologies described above is:

A non-contact electrocardiogram sensor for a wearable real-time multi-channel non-contact electrocardiometer, characterized in that:

The ECG sensor is composed of an independent 4-layer circular PCB board, including detection elements formed inside the 4-layer circular PCB board, and electrocardiographic signal amplification mounted on the top wiring layer of the 4-layer circular PCB board. module and a four-core connector, the ECG signal amplification module is connected to the detection element, and the output end of the ECG signal amplification module is connected to the electrocardiograph host through a four-core connection through a signal wire; The second layer of the 4-layer circular PCB board is all coated with copper to form a shielding layer; the third layer of the 4-layer circular PCB board is an inner wiring layer, which is used for power and signal connections inside the ECG sensor;

The detection element is formed by a multi-layer PCB process, and is composed of a detection electrode and a three-dimensional shielding cavity surrounding the detection electrode: the bottom wiring layer of the 4-layer circular PCB board is divided into a detection electrode located in the center of the PCB board and a surrounding detection electrode The shielding ring; several via holes evenly distributed along the shielding ring connect the shielding ring and the shielding layer to form a three-dimensional shielding cavity with a cage-shaped three-dimensional structure; High-impedance electrical isolation between the skins, the detection electrodes acquire ECG signals through capacitive coupling in a non-contact manner;

The detection electrode is connected to the non-inverting input terminal of the ECG signal amplification module through a DC blocking capacitor; the three-dimensional shielding cavity is connected to the reverse input terminal of the ECG signal amplification module through a resistor, so as to eliminate the impact of external interference signals on the detection electrode. Impact.

A preferred technical solution of the non-contact ECG sensor of the present invention is characterized in that the diameter of the 4-layer circular PCB board is 2.8cm, and the detection electrode is an annular copper-clad area with a diameter of 2.5cm. The shielding ring is an annular copper clad area with a width of 1.4 mm, and an insulating gap of 0.1 mm is left between the detection electrode and the shielding ring.

A better technical solution of the non-contact ECG sensor of the present invention is characterized in that the four-core connector is a 3.5mm four-core audio socket, and the output end of the ECG signal amplification module passes through the length The up to 50cm audio signal wire is connected to the host of the electrocardiograph.

Another purpose of this utility model is to provide a wearable multi-channel ECG sampling underwear using the above-mentioned non-contact ECG sensor, which solves the problem of obtaining low-cost information with medical and physiological significance, and can be used for diagnosing heart diseases and monitoring cardiac physiology. Technical issues of high-quality multi-channel ECG signals in different conditions. The technical solution adopted by the present invention to solve the problems of the technologies described above is:

A wearable multi-channel ECG sampling underwear using the above-mentioned non-contact ECG sensor, comprising at least 3 ECG sensors connected to the host of the ECG instrument, characterized in that: the ECG sensor passes through four The core coupler is connected to the electrocardiogram sampling unit of the electrocardiogram mainframe; the electrocardiogram sampling unit is formed by connecting the instrumentation operational amplifiers of each electrocardiogram signal channel with eight-channel analog-to-digital converters; the instrumentation of each electrocardiogram signal channel The operational amplifier is connected in a multi-channel differential input mode, and the multi-channel ECG signals obtained by the ECG sensor are transmitted to the signal processing unit of the host computer of the ECG instrument. The inverting input terminals of the instrumentation operational amplifiers of the independent ECG signal channels are connected in parallel; the remaining ECG sensors are used as the differential positive poles of the independent ECG signal channels, and are respectively connected to the non-inverting inputs of the instrumentation operational amplifiers of each ECG signal channel. Terminal; each ECG signal from the ECG sensor is sent to the corresponding analog input channel of the eight-channel analog-to-digital converter after being amplified by the high-power differential of the operational amplifier used in the instrument, and the ECG sampling signal is converted into a digital signal.

A preferred technical solution of the wearable multi-channel ECG sampling underwear of the present invention is characterized in that it also includes an alarm unit arranged on the middle and upper part of the chest of the elastic clothing material, and the alarm unit includes a In an emergency, the wearer can record his own voice signal by pressing the alarm button, and send the alarm information to the connected smart terminal through the host of the electrocardiograph, or send the alarm message to the connected smart terminal through the public data communication network The alarm information is automatically broadcast to the customer service center or medical emergency center.

A better technical solution of the wearable multi-channel ECG sampling underwear of the present invention is characterized in that the ECG sampling underwear is equipped with 6 ECG sensors; the ECG sampling unit includes Six instrumentation operational amplifiers connected in a multi-channel differential input mode; the electrocardiogram sensor is configured according to the conventional electrocardiogram electrode position to obtain multi-channel electrocardiogram signals similar to professional medical electrocardiographs.

The beneficial effects of the utility model are:

1. The non-contact ECG sensor of the present utility model adopts the detection element formed inside the PCB board, which greatly reduces the volume and installation area of the sensor, and makes the wearable real-time multi-channel non-contact ECG instrument in the same size More electrodes can be placed and arranged in the chest space, thereby improving the spatial resolution of ECG signal sampling.

2. The wearable multi-channel ECG sampling underwear of the present invention adopts a plurality of separate small-sized non-contact ECG sensors, and connects to the ECG instrument through a signal wire with a length of up to 50 cm in a multi-channel differential input mode. The host, in the daily electromagnetic environment where there is interference from wireless signals such as mobile phone signals and WiFi, the collected ECG signals have a signal quality comparable to the traditional measurement method of Holter, a professional medical electrocardiometer.

Description of drawings

Fig. 1 is a three-dimensional structural schematic diagram of a non-contact ECG sensor of the present invention;

Fig. 2 is a structural sectional view of the non-contact ECG sensor of the present invention;

Fig. 3 is a sectional view of the sensing surface of the non-contact ECG sensor of the present invention;

Fig. 4 is a schematic diagram of the chest configuration scheme of the wearable multi-channel ECG sampling underwear of the present invention;

Fig. 5 is a schematic diagram of the waist configuration scheme of the wearable multi-channel ECG sampling underwear of the present invention;

Fig. 6 is a schematic diagram of the shoulder configuration scheme of the wearable multi-channel ECG sampling underwear of the present invention;

Fig. 7 is the circuit schematic diagram of the wearable real-time multi-channel non-contact electrocardiograph using the electrocardiogram sensor of the present invention;

Fig. 8 is a schematic diagram of the electrocardiograph mainframe of the wearable real-time multi-channel non-contact electrocardiograph of the present invention;

Fig. 9 is a flow chart of the machine learning method of the wearable real-time multi-channel non-contact electrocardiograph of the present invention;

Fig. 10 is a flow chart of the heartbeat estimation method of the wearable real-time multi-channel non-contact electrocardiograph of the present invention.

In the figure, 1-top wiring layer, 2-ECG signal amplification module, 201-preamplifier, 202-secondary amplifier, 3-four-core connector, 4-shielding layer, 5-inner wiring layer, 6-detection Electrode, 7-shielding ring, 8-via hole, 9-solder mask insulation layer, 10-detection element, 20-ECG main unit, 21-ECG sampling unit, 211-op amp for instrumentation, 212-analog-to-digital conversion device, 22-signal processing unit, 221-first-level data cache, 222-heartbeat capture module, 223-mass storage, 224-secondary data cache, 225-Web server, 23-wireless transmission unit, 30-intelligent terminal, 40-elastic clothing, 50-alarm unit, Hs-ECG sensors H1-H6.

detailed description

In order to better understand the above-mentioned technical solution of the utility model, a further detailed description will be given below in conjunction with the accompanying drawings and embodiments.

A group of embodiments of the wearable real-time multi-channel non-contact electrocardiometer of the present utility model are shown in Figures 4 to 7, and are composed of an electrocardiometer host 20 and k independent small-sized non-contact electrocardiographic sensors Hs, Wherein s=1～k, 9≥k≥3, the described ECG sensor Hs is connected to the ECG host 20 in a multi-channel differential input mode through a signal wire whose length can reach 50 cm; each ECG sensor Hs is composed of An independent 4-layer circular PCB with a diameter of 2.8cm is formed; in the three configuration embodiments of Fig. 4 to Fig. 6, k=6, and 6 electrocardiographic sensors H1-H6 are arranged on the chest of the elastic clothing material 40; In the three configuration embodiments shown in FIG. 4 to FIG. 6 , the ECG sensors H1 to H6 are distributed at conventional ECG electrode positions to obtain multi-channel ECG signals similar to a professional medical electrocardiometer (Holter).

An embodiment of the non-contact ECG sensor used for the present utility model is shown in Figures 1 to 3, including a detection element 10 formed inside a 4-layer circular PCB board, and mounted on the top layer of a 4-layer circular PCB board for wiring Layer ECG signal amplification module 2 and four-core connector 3; the ECG signal amplification module 2 is connected to the detection element 10, and the four-core connector 3 connects the ECG sensor Hs to the ECG The instrument host 20; the ECG signal amplification module 2 and the four-core connector 3 are mounted on the top layer wiring layer 1 of the 4-layer circular PCB board; the second layer of the 4-layer circular PCB board is all formed by copper coating Shielding layer 4; the third layer of the 4-layer circular PCB board is an inner wiring layer 5, which is used for power supply and signal connection inside the ECG sensor;

The detection element 10 is formed by a multi-layer PCB process, and is composed of a detection electrode 6 and a three-dimensional shielding cavity surrounding the detection electrode 6: the bottom wiring layer of the 4-layer circular PCB board is divided into a detection electrode 6 and a shielding ring 7, The detection electrode 6 is a circular copper-clad area with a diameter of 2.5 cm, located at the center of a 4-layer circular PCB; the shielding ring 7 is an annular copper-clad area with a width of 1.4 mm, located at the outer circle of the circular PCB; An insulation gap of 0.1 mm is left between the detection electrode 6 and the shielding ring 7; 40 via holes 8 are evenly distributed along the shielding ring 7 to connect the shielding ring 7 and the shielding layer 4 to form a three-dimensional shielding cavity with a cage-shaped three-dimensional structure ; According to a preferred embodiment of the ECG sensor of the present utility model, the via hole 8 is a metallized hole with a diameter of 3mil; The high-impedance electric isolation between them, its insulation resistance> 1G ohm, described detecting electrode 6 obtains electrocardiogram signal through capacitive coupling in non-contact mode; There is no direct electrical connection between them, unlike the existing wet electrodes, which need to rely on the skin surface to apply gel to ensure good conduction, nor like the existing non-contact ECG sensors with bare induction strips, which need to rely on the user to wear underwear or T-shirt to achieve relative "insulation".

The electrocardiographic signal amplification module 2 includes a preamplifier 201 and a secondary amplifier 202, and the detection electrode 6 is connected to the non-inverting input end of the preamplifier 201 through a DC blocking capacitor; the three-dimensional shielding cavity is passed through a 10K ohm The resistor is connected to the inverting input end of the preamplifier 201 to eliminate the influence of external interference signals on the detection electrodes 6; the preamplifier 201 is connected to the secondary amplifier 202 through an electrocardiographic signal filter. In a typical embodiment of the ECG sensor of the present invention, the ECG signal filter is composed of a 0.5Hz high-pass filter, a 100Hz low-pass filter and a 50Hz trap circuit. Considering that the ECG signal is a low-frequency signal, according to the preferred embodiment shown in Figure 1, the ECG sensor of the present utility model uses a 3.5mm four-core audio socket as the four-core connector 3, replacing the miniUSB used for electrode connection in the technical solution of CN102657524B interface. Since the centrally symmetrical 3.5mm four-core audio socket in the shape of a cylinder has no direction, the cable plug can be inserted into the socket from any angle of 360 degrees, which is more convenient to use than the miniUSB interface that must be inserted into the socket from a fixed angle. The output end of the electrocardiographic signal amplification module 2 is connected to the electrocardiograph host through a four-core connection 3 through an audio signal wire whose length can reach 50 cm. In order to ensure the best wearing comfort for users in different postures, the utility model provides three sampling system distribution schemes, the chest scheme shown in Figure 4 is suitable for sports, and the waist scheme shown in Figure 5 is suitable for comfort requirements For higher light activities, the shoulder scheme shown in Figure 6 is suitable for sleep, and users can choose and configure it at will according to their preferences.

Since the ECG sensor of the utility model adopts a three-dimensional shielding cavity structure and a multi-channel differential input mode, it can effectively eliminate external interference and significantly improve the signal-to-noise ratio. The ECG sensor of the utility model reduces the diameter of the circular PCB board to 2.8cm Compared with the 3.9cm circular PCB board with a diameter of 3.9cm in the CN102657524B technical solution, the electrocardiographic sensor area of the utility model is reduced by 48.5%. By greatly reducing the volume and installation area of the non-contact ECG sensor, the wearable real-time multi-channel non-contact ECG instrument of the present invention can place and arrange more electrodes in the chest space of the same size to improve Spatial resolution of ECG signal sampling. As shown in Figures 4-6, the ECG sensor of the present invention can be easily built into elastic clothing fabrics, and distributed in classic positions to obtain Holter-like ECG signals.

An embodiment of the wearable real-time multi-channel non-contact electrocardiograph of the present utility model is shown in FIG. Described electrocardiogram sampling unit 21 is made of k-1 operational amplifiers 211 and eight-channel analog-to-digital converters 212 connected by the multi-channel differential input mode of each electrocardiographic signal channel; k-1 operational amplifiers 211 for instruments It is connected in a multi-channel differential input mode, wherein the first ECG sensor H1 is used as a common differential negative pole, and is connected in parallel with the inverting input terminal of the instrument operational amplifier 211 of each independent ECG signal channel; the rest of the ECG sensors H2～Hk, as the differential positive poles of k-1 independent ECG signal channels, are respectively connected to the non-inverting input terminals of the instrumentation operational amplifier 211 of each ECG signal channel; each ECG signal from the ECG sensor Hs, through the instrumentation After the operational amplifier 211 has a high power (amplification factor > 280 times) differential amplification, it is sent to the corresponding analog input channel of the eight-channel analog-to-digital converter 212, and the sampling is converted into a 16-bit digital signal, which is sent to the signal processing unit 22 as ECG sampling data.

Since the input impedance of the non-contact electrode is very high (G ohm level), it is very sensitive to environmental noise, which can easily cause the saturation of the back-end signal conditioning circuit. The instrumentation operational amplifier 211 of this embodiment is the INA333 instrumentation differential operational amplifier of Texas Instruments. The INA333 has a very high common-mode rejection ratio and extremely low input offset voltage, and can effectively filter out ECG from non-contact electrodes. The common mode interference mixed in the signal fundamentally improves the signal-to-noise ratio of the input ECG signal and improves the signal quality.

According to the embodiment of the electrocardiograph host 20 of the wearable real-time multi-channel non-contact electrocardiograph shown in Fig. 8, the described signal processing unit 22 includes a primary data cache 221, a heartbeat capture module 222, and a mass memory 223, the secondary data cache 224 and the Web server 225; the ECG sampling unit 21 transmits the ECG sampling data obtained by sampling conversion to the primary data cache 221, and stores it in the mass storage 223 as original data; The heartbeat capture module 222 reads the ECG sampling data from the primary data cache 221, calculates the heartbeat quantization estimated value and judges and captures the ECG signal according to the heartbeat quantization estimated value, and transmits the ECG signal data to the secondary data cache 224; The Web server 225 reads the ECG signal data from the secondary data cache 224, converts it into an HTML standard format, and provides an online ECG signal display function through the wireless transmission unit 23; The ECG signal data is converted into an HTML standard format to provide historical heartbeat data playback and original ECG sampling data download functions through the wireless transmission unit 23 .

According to an embodiment of the electrocardiograph host 20 of the wearable real-time multi-channel non-contact electrocardiograph of the present invention, the high-capacity memory 223 adopts a highly integrated non-volatile memory chip, and its available storage capacity At least 2GB, which can meet the requirements of long-term uninterrupted storage of high-precision high-sampling rate multi-channel ECG signals; taking 16bit high-precision ECG signals as an example, if the sampling rate is 1KHz per channel, 24-hour uninterrupted storage must be achieved8 For channel sampling data, the storage capacity needs to be at least 16*1000/8*3600*24*8=1382400000byte≈1.4GB; the technical solution of the utility model realizes that ECG sampling storage does not depend on external wireless signal coverage; no matter where the user is Whether the location is covered by wireless signals, the ECG data will not be lost.

The wireless transmission unit 23 includes or Bluetooth, ZigBee or WiFi dedicated wireless network, or 2G, 3G or 4G public network; the intelligent terminal 30 only needs to access the Web server 225 through the wireless transmission unit 23 based on the HTML protocol, without installing any special software Or plug-in just can realize the cross-platform display of electrocardiogram data; Described intelligent terminal 30 is the cross-platform terminal equipment that configures HTML standard web browser, comprises based on any operating system platform of Windows, Linux, Unix, iOS or Android personal computer, laptop, smart phone, intelligent mobile terminal, or a monitoring terminal of a telemedicine emergency center; the web browser includes IE, Chrome, Firefox, and Safari browsers that support HTML standards, or an application app based on web services.

According to the embodiment of the configuration scheme shown in Figures 4-6, the wearable real-time multi-channel non-contact electrocardiograph of the present invention also includes an alarm unit 50 connected to the electrocardiometer host 20, and the alarm unit 50 is configured In the upper middle part of the chest of the elastic clothing material 40, a miniature microphone and an alarm button are included; , such as pressing for more than 3 seconds) to record your own voice signal (such as "I feel flustered and chest tight"), and send an alarm to the connected smart terminal 30 (handheld devices such as mobile phones). After receiving the alarm, the terminal 30 can automatically broadcast to the customer service center or medical emergency center through the public data communication network in time; the user or the medical staff can receive the alarm from the wearer's massive historical ECG data according to the wearer's alarm time and voice information. Quickly and accurately extract ECG data fragments at important time points and correlate them with the wearer's physical feelings, realize efficient data search, and capture ECG data of sporadic and random cardiac abnormalities, premature beats or arrhythmias; in non-emergency Under circumstances (such as running, walking, eating, taking medicine, etc.), the wearer clicks or short presses the alarm button, and the alarm unit 50 can also be used for voice recording of the event or the wearer’s state, for later data analysis and disease Diagnostic collection data.

The ECG monitoring method used for the above-mentioned wearable real-time multi-channel non-contact ECG instrument of the present utility model comprises the following steps:

S10: Obtain n pieces of historical ECG sampling data x _i from the large-capacity memory 223 of the electrocardiograph host 20, where i=1...n, x _i =[ _xi,1 x _i,2 ... _{xi ,m} ] is the i-th sample vector, m is the maximum dimension of the sample vector x _i , n is the total sample size and n is much larger than m;

S20: During the initial development and phased update of the product, use the original ECG sampling data _xi to establish a sample matrix in the development environment and execute the machine learning program, and obtain the target vector β with the minimum error value through a limited number of iterations as a constant parameter Stored in the memory of the electrocardiograph host 20, it is used to capture the weight vector of any individual's heartbeat at any time; this step has a large amount of calculation and takes a long time, and it is only used on the computing platform of the development environment during the initial development and phased update of the product implement.

S30: The electrocardiograph host 20 acquires real-time sampled data x _i through electrocardiographic sampling, and uses the pre-stored target vector β to calculate the quantified estimated value of heartbeat Heartbeat estimation is performed on the real-time sampled data x _i to realize fast heartbeat capture, where T is a matrix transposition operation. Since f( _xi , β) mainly includes linear operations, the amount of calculation is very small, so the speed of the heartbeat capture algorithm in this step is very fast, and the CPU occupation is very small. The ability to complete a heartbeat estimate every 50-100 milliseconds, can capture ECG signals using simple and fast linear operations while fully maintaining the original signal waveform.

According to the embodiment of the heartbeat capture method of the present invention shown in FIG. 9, the step S20 executes the machine learning program according to the following steps:

S200: Generate a random value vector with the same dimension as the target vector β as the initial value of β; according to a preferred embodiment, the initial value of the target vector β is based on a random value vector between 0 and 1 (same as β Dimensions), and then multiply the value of each dimension of the random value vector by 0.2 and then subtract 0.1 to generate.

S210: Establish a sample matrix ( _xi , y _i )i=1...n in the development environment, where y _i is the judgment value of whether the historical ECG sampling data _xi is a heartbeat, namely

S220: Establish an error function J(β) based on the sample matrix (x _i , y _i ):

$\mathrm{<span\; class="notranslate">\; J</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}<span\; class="notranslate">\; \&lsqb;</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; log</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; log</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&rsqb;</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; no</span>}}\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; b</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>$

in, j is the dimension of the target vector β, the constant λ is used to control the size of the iteration increment, λ≈1;

S230: Establish a gradient function dJ/dβ based on the sample matrix (x _i , y _i ):

When j = 0,

When j≥1;

Because the target vector β is multidimensional (the total dimension is m, m>1), its corresponding gradient function is also multidimensional, that is, each dimension of the gradient function (such as the jth dimension) is the error function J(β) based on the target The first partial derivative of the corresponding dimension of the vector. At the same time, in the iterative calculation process, a dimension with a constant of 1 is artificially added to the sample vector _xi , marked as x ₀ , and a dimension b ₀ corresponding to x ₀ is added to the target vector β accordingly, which is used in the general linear An estimate of the intercept in a regression. It should be emphasized that b ₀ is only used in the iterative calculation for the target vector β. When the calculation is finished and the best target vector based on all samples is found, b ₀ will be removed.

S240: Taking error function J(β) and gradient function dJ/dβ as input, calling GNCfminunc function, performing iterative operation to update target vector β:

S250: Find the local extremum through a finite number of iterative algorithms, and finally obtain the target vector β=[b ₁ b ₂ ...b _m ] to obtain the minimum error value, and send it to the electrocardiograph host 20 as a constant parameter and store it in it in memory.

According to the embodiment of the heartbeat capture method of the present invention shown in FIG. 10, the step S30 performs heartbeat estimation according to the following steps:

S300: Obtain the real-time sampling data _xi of the ECG sensor from the first-level data cache, where _xi is an array with a length of several hundred bits, which contains multi-channel sampling data with a length of 50-100 milliseconds;

S310: Using the pre-stored target vector β, calculate the heartbeat quantization estimated value of the real-time sampled data x _i The calculation result of y′ _i is between 0 and 1;

S320: Determine whether the heartbeat quantization estimated value y′ _i is greater than 0.5, if y′ _i >0.5, go to step S330, otherwise go to step S340;

S330: Determine that the sampling data x _i is a real heartbeat, send the heartbeat time into the secondary data cache, and return to step S300 to wait for the next ECG sampling;

S340: Determine that the sampling data does not contain valid heartbeat data, and return to step S300 to wait for the next ECG sampling.

In order to embody the core utility model points of the present invention, the above-mentioned heartbeat capture algorithm only describes the basic algorithm structure in detail. Although in the above embodiments, the target vector β is a 1×m array, the heartbeat capture method of the present invention is also applicable to more complex mathematical models, for example, for a multi-layer neural network, when calculating f( _xi , In the process of β), it is necessary to add a matrix describing the hidden layer, so there may be multiple target vectors involved, that is, multiple operations need to be performed sequentially starting from _xi , each layer performs a matrix multiplication operation, and finally obtains a sample-based x _i , heartbeat estimation result f( _xi ,β) whose value is between 0 and 1. The specific operation steps are different due to different mathematical models, but the basic structure and idea of the heartbeat capture algorithm are the same.

Those of ordinary skill in the technical field should recognize that the above embodiments are only used to illustrate the technical solutions of the present utility model, and are not used as limitations to the present utility model. The changes and modifications made in the embodiments will all fall within the protection scope of the claims of the present utility model.

Claims (6)

1. A non-contact electrocardiographic sensor for wearable real-time multi-channel non-contact electrocardiography, characterized in that: the electrocardiographic sensor is made of an independent 4-layer circular PCB board, included in The detection element formed inside the 4-layer circular PCB board, and the ECG signal amplification module and the four-core connector mounted on the top wiring layer of the 4-layer circular PCB board, the ECG signal amplification module is connected to the The detection element, the output end of the electrocardiographic signal amplification module, is connected to the electrocardiograph mainframe via a four-core connection through a signal wire; the second layer of the 4-layer circular PCB board is all coated with copper to form a shielding layer; The third layer of the 4-layer circular PCB board is the inner wiring layer, which is used for the power supply and signal connection inside the ECG sensor; The detection element is formed by a multi-layer PCB process, and is composed of a detection electrode and a three-dimensional shielding cavity surrounding the detection electrode: the bottom wiring layer of the 4-layer circular PCB board is divided into a detection electrode located in the center of the PCB board and a surrounding detection electrode The shielding ring; several via holes evenly distributed along the shielding ring connect the shielding ring and the shielding layer to form a three-dimensional shielding cavity with a cage-shaped three-dimensional structure; High-impedance electrical isolation between the skins, the detection electrodes acquire ECG signals through capacitive coupling in a non-contact manner; The detection electrode is connected to the non-inverting input terminal of the ECG signal amplification module through a DC blocking capacitor; the three-dimensional shielding cavity is connected to the reverse input terminal of the ECG signal amplification module through a resistor, so as to eliminate the impact of external interference signals on the detection electrode. Impact. 2. The non-contact ECG sensor according to claim 1, characterized in that the diameter of the 4-layer circular PCB board is 2.8cm, the detection electrode is an annular copper-clad area with a diameter of 2.5cm, and the shielding The ring is an annular copper clad area with a width of 1.4mm, and an insulating gap of 0.1mm is left between the detection electrode and the shielding ring. 3. The non-contact ECG sensor according to claim 1, characterized in that the four-core connector is a 3.5mm four-core audio socket, and the output end of the ECG signal amplification module can reach a length of 50 cm. The audio signal lead is connected to the electrocardiograph host. 4. A wearable multi-channel electrocardiogram sampling underwear using the non-contact electrocardiogram sensor described in claim 1, 2 or 3, comprising at least 3 electrocardiogram sensors that are connected to the electrocardiograph host, and It is characterized in that: the ECG sensor is connected to the ECG sampling unit of the electrocardiograph host through a four-core connector; the ECG sampling unit is composed of an instrumental operational amplifier for each ECG signal channel and an eight-channel analog-to-digital conversion The operational amplifiers of each ECG signal channel are connected in a multi-channel differential input mode, and the multi-channel ECG signals obtained by the ECG sensor are transmitted to the signal processing unit of the host computer of the ECG instrument. Among them, the first The ECG sensor is used as a common differential negative pole, and is connected in parallel with the inverting input terminal of the instrumental operational amplifier of each independent ECG signal channel; the other ECG sensors are used as differential positive poles of each independent ECG signal channel, respectively connected to each heart The non-inverting input terminal of the operational amplifier of the electrical signal channel; each ECG signal from the ECG sensor is sent to the corresponding analog input channel of the eight-channel analog-to-digital converter after being amplified by the high-power differential of the operational amplifier for the instrument, and the The ECG sampling signal is converted into a digital signal. 5. The wearable multi-channel electrocardiographic sampling underwear according to claim 4, is characterized in that it also includes an alarm unit configured on the middle and upper part of the chest of the elastic clothing material, and the described alarm unit includes a Miniature microphone and alarm button. In an emergency, the wearer can record his own voice signal by pressing the alarm button, and send the alarm information to the connected smart terminal through the electrocardiograph host, or automatically send the alarm information through the public data communication network. Broadcast to customer service center or medical emergency center. 6. The wearable multi-channel ECG sampling underwear according to claim 4, characterized in that the ECG sampling underwear is equipped with 6 ECG sensors; the ECG sampling unit includes 6 or more An instrumentation operational amplifier connected in a multi-channel differential input mode; the electrocardiogram sensor is configured in accordance with conventional electrocardiogram electrode positions to obtain multi-channel electrocardiogram signals similar to professional medical electrocardiographs.