Pulse wave sensor and manufacturing method thereof, and blood pressure monitoring device and method

Abstract

The invention provides a high-precision flexible pulse wave sensor, a manufacturing method of the high-precision flexible pulse wave sensor and a blood pressure monitoring device and method based on pulse wave signals. The sensor is prepared by taking melamine sponge, polydopamine (PDA) and carbon nanotubes (CNT) as main materials through specific soaking and drying processes, and has good linear response and ultralow detection limit. In combination with an extreme learning machine (ELM) blood pressure model, the device can realize high-precision and real-time blood pressure prediction, improves the prediction precision and speed, is suitable for stably measuring pulse waves for a long time, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to the technical field of real-time pressure-free blood pressure monitoring, and in particular to a pulse wave sensor and a manufacturing method thereof, and a blood pressure monitoring device and method. Background Art

[0002] Hypertension, as a common chronic disease in people's daily lives, has become one of the most important risk factors for cardiovascular disease. The "Summary of China's Cardiovascular Health and Disease Report 2021" released in 2022 shows that in 2019, deaths caused by CVD in my country ranked first among major disease mortality rates, among which patients with hypertension accounted for as high as three-quarters, with the highest prevalence; in addition, the awareness rate of patients with hypertension is very low, even less than 10% among adults under 25 years old, which is very unfavorable for the treatment of CVD. And diseases such as sudden death caused by high blood pressure are usually sudden, so daily continuous blood pressure monitoring is very important.

[0003] Non-invasive continuous blood pressure measurement has the following important significance: First, it can predict and monitor heart disease, prevent sudden CVD, assist doctors in diagnosis and treatment, help patients use drugs rationally, and reduce mortality; second, the measurement of hypertension in the clinic has the white coat effect, usually because the patient is nervous, which will cause the blood pressure measurement value to be high. Continuous measurement can truly reflect the patient's blood pressure level and eliminate the randomness of the measurement; current continuous blood pressure monitoring usually adopts an invasive arterial method, which will cause many complications and is only suitable for patients in intensive care units. Non-invasive continuous blood pressure measurement can avoid these problems.

[0004] There are four common non-invasive blood pressure measurement methods. The constant volume method and the arterial tension method both require precise external pressure to fix the state of the blood vessels, which is difficult to operate and will cause local ischemia; the pulse wave characteristic parameter method has high requirements on the quality of the pulse wave signal, and some data processing can easily distort the original pulse wave signal, so few people use this method; and the pulse wave velocity method has huge advantages over the other three methods. The puncture method is to directly measure the blood pressure by directing the blood flow out of the body. This is measured according to the definition of blood pressure and is the gold standard, but it is invasive and is usually only used in intensive care units. Auscultation and oscillometric methods are commonly used clinically. Auscultation is the gold standard for non-invasive blood pressure measurement, but it is more complicated to operate and requires professionals to operate. Therefore, the oscillometric method has been widely used in recent years, but both methods require complete blockage of arterial blood flow, which will cause discomfort to patients, so it is not suitable for long-term measurement. It can be seen that the pulse wave velocity method uses a pressure sensor to sense the pulse wave signal without compressing the artery. Compared with several common blood pressure measurement methods, it has obvious advantages in daily wear, stability and continuity. Therefore, the pulse wave velocity method has attracted widespread attention due to its comfort, accuracy and continuity, and is of great significance in the field of blood pressure monitoring.

[0005] Flexible pressure sensors have been a hot topic of research in recent years. They have been widely used in gesture recognition, health monitoring and other fields. The principle of pressure sensors is that after the compression layer is strained by stress changes, the performance of the material changes accordingly. According to different materials and different signals, sensors can be divided into several types: capacitance, inductance, piezoelectric, and triboelectric. As the name suggests, resistance and capacitance are changes in these two signals. The principle of piezoelectric is based on the piezoelectric effect, and triboelectric mainly uses friction to generate electricity. Among them, resistive pressure sensors have great advantages due to their simple structure, strong durability, high signal stability and simple detection method. However, resistive pressure sensors have problems such as low sensitivity, small linear response range, and long response time, and there is a lot of room for improvement. How to prepare high-performance sensors for detecting human physiological signals such as heart rate and pulse waves is also an urgent problem to be solved.

[0006] At present, blood pressure collection devices are mainly divided into two categories, one is the mercury column sphygmomanometer based on Korotkoff sound auscultation method, and the other is the cuff-type electronic pressure sphygmomanometer. The cuff design of these two sphygmomanometers is clumsy and heavy, which cannot meet the needs of continuous wearable real-time blood pressure detection, and this method may cause discomfort to the tester.

[0007] It should be noted that the information disclosed in the above background technology section is only used for understanding the background of the present application, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the invention

[0008] The main purpose of the present invention is to overcome the defects existing in the above-mentioned background technology and to provide a pulse wave sensor and a manufacturing method thereof, and a blood pressure monitoring device and method.

[0009] To achieve the above object, the present invention adopts the following technical solutions:

[0010] A pulse wave sensor, comprising:

[0011] Melamine sponge;

[0012] A polydopamine (PDA) layer, covering the melamine sponge, for forming a conductive layer;

[0013] Carbon nanotubes (CNTs), dispersed in the polydopamine (PDA) layer to enhance conductivity and sensitivity;

[0014] A conductive circuit, comprising a polyimide copper film and interdigital electrodes formed on the polyimide copper film; wherein the interdigital electrodes cooperate with the polydopamine (PDA) layer and carbon nanotubes (CNT) to transmit electrical signals;

[0015] The upper and lower PDMS films encapsulate the melamine sponge, polydopamine (PDA) layer, carbon nanotubes (CNT) and conductive circuits to provide flexibility and protection.

[0016] Furthermore, the PDMS film is plasma treated to enhance adhesion.

[0017] Furthermore, the width of the interdigitated electrode channel should be less than 1 cm, and the width of a single electrode should be greater than 20 microns.

[0018] A method for manufacturing a pulse wave sensor comprises the following steps:

[0019] S1. Prepare melamine sponge;

[0020] S2, covering the melamine sponge with a polydopamine (PDA) layer to form a conductive layer;

[0021] S3, dispersing carbon nanotubes (CNT) in the polydopamine (PDA) layer to enhance conductivity and sensitivity;

[0022] S4, preparing a conductive circuit, including forming interdigital electrodes on the polyimide copper film;

[0023] S5. Encapsulate the melamine sponge, polydopamine (PDA) layer, carbon nanotube (CNT) and conductive circuit between upper and lower PDMS films to provide flexibility and protection; wherein the interdigitated electrodes cooperate with the polydopamine (PDA) layer and carbon nanotube (CNT) to transmit electrical signals.

[0024] Furthermore, in step S2, the formation of the polydopamine (PDA) layer comprises adding Tris-HCl buffer to dopamine hydrochloride, and then soaking the sponge cleaned in step S1 in the solution for a soaking time greater than 1 hour.

[0025] Furthermore, in step S3, multi-walled carbon nanotubes (CNTs) are poured into a container, isopropanol is dropped into it and ultrasonic treatment is performed to fully disperse the CNTs in the isopropanol, the mass fraction of CNTs in the dispersion is 0.1wt% to 10wt%, and the melamine sponge covered with a polydopamine (PDA) layer is immersed in the CNT dispersion and then dried.

[0026] Furthermore, the immersion time in the dispersion is not less than 10 minutes, the drying temperature is 40° C.-100° C., and the drying time is not less than 1 minute.

[0027] Furthermore, in step S5, at least one side of the upper and lower PDMS films is treated with plasma, and then the two PDMS films are pressure-bonded.

[0028] A blood pressure monitoring device based on a pulse wave signal comprises the pulse wave sensor and a data processing system.

[0029] A blood pressure monitoring method based on pulse wave transit time comprises the following steps:

[0030] S1, using the pulse wave sensor to collect a human body's pulse wave signal;

[0031] S2, preprocessing the collected pulse wave signal to extract the pulse wave transit time (PTT), pulse wave velocity (PWV) and pulse wave main peak value (MX);

[0032] S3, using a multimodal blood pressure parameter model to calculate the first-level blood pressure value, the model is BP = aPWV^2 + bPWV + cln(PWV) + d, where a, b, c, d are model parameters calculated by nonlinear regression of experimental data;

[0033] S4, taking the first-level blood pressure value obtained in step S3 and the MX obtained in step S2 as input feature parameters, and importing them into the trained extreme learning machine (ELM) blood pressure model;

[0034] S5, normalizing MX in the ELM model, and combining the normalized MX with other physiological parameters into an input matrix H;

[0035] S6. Set the number of hidden layer nodes of ELM to 1, randomly generate hidden layer weights W and hidden layer biases B, and calculate the hidden layer output matrix G, which is expressed as H = σ(WX+B), where σ is the activation function;

[0036] S7. Once the output H of the hidden layer is determined, ELM uses the least squares method to solve the output weight β, so that the predicted value Y' of the output layer is close to the true target value Y, and the formula is expressed as β = (H T H+λI) -1 H T Y, where λ is the regularization parameter and I is the identity matrix;

[0037] S8. Use the output weight β obtained in S7 to predict the experimental sample data X, and the formula is expressed as Y'=σ(WX+B)β.

[0038] The present invention has the following beneficial effects:

[0039] The present invention provides a pulse wave sensor and a manufacturing method thereof, a blood pressure monitoring device and a method. The pulse wave sensor is an ultra-thin flexible resistive pressure sensor. The pulse wave sensor has good linear response (sensitivity 6.23 kPa-1) and ultra-low detection limit (less than 100 Pa) in a wide pressure range (0-350 kPa). The electrical response is still stable after 1500 cycles of compression, and the response time is very short (100 ms).

[0040] After the pulse wave signal is collected by the sensor, the acquisition module obtains and calculates the pulse wave transit time (PTT), the pulse wave velocity (PWV) and the pulse wave main peak value (MX). The signal processing module uses the blood pressure parameter model to calculate the first-level blood pressure value, and then combines it with MX and inputs it into the trained extreme learning machine (ELM) blood pressure model to obtain the final blood pressure prediction value. The sensor has a good linear response (sensitivity 6.23kPa-1) and an ultra-low detection limit (less than 100Pa) in a wide pressure range (0-350kPa). The electrical response remains stable after cyclic compression, and has a very short response time (100ms). The device combines the extreme learning machine (ELM) blood pressure model in the present invention to achieve high-precision and real-time blood pressure prediction, improve the accuracy and speed of blood pressure prediction, and realize continuous real-time blood pressure measurement. The sensor solves the current situation that pulse wave detection depends on pressurization, and high-quality pulse wave signals can be obtained when placed at the carotid artery and radial artery. While developing an ultra-light and convenient blood pressure monitoring method that does not require pressurization, the present invention improves the robustness and reliability of the detection system, and realizes a high-precision real-time continuous blood pressure monitoring method. The construction process is simple, the cost is low, and there is no need to use too many experimental instruments. It has wide applicability and can stably measure pulse waves for a long time, showing that it has great application prospects in wearable pulse wave detection.

[0041] Compared with the prior art, the present invention has the following outstanding technical effects and advantages:

[0042] The present invention provides a blood pressure monitoring device based on a pulse wave signal, designs a device based on a resistive flexible pressure sensor that can simultaneously measure two pulse wave signals, and uses the blood pressure monitoring device to non-invasively and continuously measure blood pressure through a pulse wave velocity method based on the relationship between the pulse wave velocity and blood pressure. At the same time, the present invention proposes a blood pressure detection method based on the pulse wave transmission time, extracts feature points from the collected pulse wave signal, and adopts a PWV multi-modulus blood pressure prediction model calculation method in the first-level blood pressure calculation. This method has further improved the calculation accuracy compared with the original PWV linear model, quadratic term model, logarithmic model, etc.; in terms of precise blood pressure measurement, the learning method of the extreme learning machine is adopted, which greatly improves the calculation accuracy. In addition, the operation method of the blood pressure detection device is simple, the detection process is comfortable, and continuous and real-time blood pressure detection can be achieved.

[0043] The invention is simple to operate, the materials are readily available and the cost is extremely low, it is ultra-light and ultra-thin, has strong biocompatibility, is flexible and stretchable, comfortable to wear and does not require pressurization, has strong robustness, and the overall device is the size of a band-aid, with great innovation and advantages in appearance and wearing comfort. In terms of blood pressure measurement, a multi-modulus blood pressure model prediction is used to ensure that the first-level blood pressure prediction value is more accurate. Secondly, the pulse wave peak value is introduced as reference data. After being imported into the trained extreme learning machine, the first-level blood pressure prediction value can be further corrected and optimized to obtain a blood pressure prediction value with small error and high accuracy.

[0044] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The figure is a flow chart of sensor preparation according to an embodiment of the present invention.

[0046] Figure 2 This is a physical picture of the sensor according to an embodiment of the present invention.

[0047] Figure 3 Schematic diagram of a blood pressure monitoring device based on a pulse wave signal according to an embodiment of the present invention.

[0048] Figure 4 This is a sensor sensitivity test diagram of an embodiment of the present invention.

[0049] Figure 5 This is a sensor response time test diagram of an embodiment of the present invention.

[0050] Figure 6 It is a signal diagram of the sensor according to an embodiment of the present invention changing with the applied pressure.

[0051] Figure 7 This is a pulse wave test diagram according to an embodiment of the present invention.

[0052] Figure 8 Schematic diagram of the overall flow of the blood pressure monitoring method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope and application of the present invention.

[0054] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element. In addition, connection can be used for fixing as well as for coupling or communication.

[0055] It should be understood that the orientation or position relationship indicated by terms such as "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0056] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0057] See also Figure 1 The embodiment of the present invention provides a pulse wave sensor, comprising: a melamine sponge 1; a polydopamine (PDA) layer, covering the melamine sponge 1, for forming a conductive layer; carbon nanotubes (CNTs), dispersed in the polydopamine (PDA) layer to enhance conductivity and sensitivity; a conductive circuit, comprising a polyimide copper film 2 and interdigital electrodes 3 formed on the polyimide copper film; wherein the interdigital electrodes 3 cooperate with the polydopamine (PDA) layer and the carbon nanotubes (CNTs) to transmit electrical signals; and upper and lower PDMS films 4, encapsulating the melamine sponge, the polydopamine (PDA) layer, the carbon nanotubes (CNTs) and the conductive circuit to provide flexibility and protection.

[0058] like Figure 1 As shown, the embodiment of the present invention also provides a method for manufacturing the pulse wave sensor, comprising the following steps:

[0059] S1. Prepare melamine sponge 1;

[0060] S2, covering the melamine sponge with a polydopamine (PDA) layer to form a conductive layer;

[0061] S3, dispersing carbon nanotubes (CNT) in the polydopamine (PDA) layer to enhance conductivity and sensitivity; obtaining a melamine sponge-based composite PDA-CNT@SPG;

[0062] S4, preparing a conductive circuit, including forming interdigital electrodes 3 on the polyimide copper film 2;

[0063] S5, encapsulating the melamine sponge, the polydopamine (PDA) layer and the carbon nanotube (CNT) (composite PDA-CNT@SPG) and the conductive circuit between the upper and lower PDMS films 4 to provide flexibility and protection; wherein the interdigital electrodes cooperate with the polydopamine (PDA) layer and the carbon nanotube (CNT) to transmit electrical signals. The interdigital electrodes 3 on the conductive circuit can be led out with wires 5.

[0064] In some specific embodiments, melamine sponge, polydopamine (PDA) and carbon nanotubes (CNT) are used as main materials, and an ultra-light conductive sponge with a mass of only 3.3 mg is obtained through simple soaking and drying. It is encapsulated with two layers of transparent PDMS film to obtain a flexible resistive pressure sensor with a thickness of only 290 μm.

[0065] The sensor has good linear response (sensitivity 6.23kPa-1) and ultra-low detection limit (less than 100Pa) in a wide pressure range (0-350kPa). The electrical response remains stable after 1500 cycles of compression, and has a very short response time (100ms).

[0066] See also Figure 3 , a blood pressure monitoring device based on pulse wave signal, including the pulse wave sensor and data processing system, which are used for resistive pressure sensing and back-end signal processing respectively.

[0067] See also Figure 8 , a blood pressure monitoring method based on pulse wave transit time, comprising the following steps:

[0068] S1, using the pulse wave sensor to collect a human body's pulse wave signal;

[0069] S2, preprocessing the collected pulse wave signal to extract the pulse wave transit time (PTT), pulse wave velocity (PWV) and pulse wave main peak value (MX);

[0070] S3, using a multimodal blood pressure parameter model to calculate the first-level blood pressure value, the model is BP = aPWV^2 + bPWV + cln(PWV) + d, where a, b, c, d are model parameters calculated by nonlinear regression of experimental data;

[0071] S4, taking the first-level blood pressure value obtained in step S3 and the MX obtained in step S2 as input feature parameters, and importing them into the trained extreme learning machine (ELM) blood pressure model;

[0072] S5, normalizing MX in the ELM model, and combining the normalized MX with other physiological parameters into an input matrix H;

[0073] S6. Set the number of hidden layer nodes of ELM to 1, randomly generate hidden layer weights W and hidden layer biases B, and calculate the hidden layer output matrix G, which is expressed as H = σ(WX+B), where σ is the activation function;

[0074] S7. Once the output H of the hidden layer is determined, ELM uses the least squares method to solve the output weight β, so that the predicted value Y' of the output layer is close to the true target value Y, and the formula is expressed as β = (H T H+λI) -1 H T Y, where λ is the regularization parameter and I is the identity matrix;

[0075] S8. Use the output weight β obtained in S7 to predict the experimental sample data X, and the formula is expressed as Y'=σ(WX+B)β.

[0076] The present invention utilizes an ultra-thin, flexible resistive pressure sensor to achieve high-precision and real-time blood pressure monitoring. The sensor uses melamine sponge, polydopamine (PDA) and carbon nanotubes (CNT) as the main materials, prepares an ultra-light conductive sponge through a simple process, and encapsulates it with two layers of transparent PDMS film to ensure the flexibility and protection of the sensor. It exhibits good linear response and ultra-low detection limit in a wide pressure range, and the electrical response is still stable even after 1500 cycles of compression, with a response time of only 100ms. Combined with the extreme learning machine (ELM) blood pressure model, the blood pressure monitoring method of the present invention not only improves the accuracy and speed of blood pressure prediction, but also realizes continuous real-time monitoring of blood pressure through the extraction and analysis of pulse wave transit time (PTT), pulse wave velocity (PWV) and pulse wave main peak peak (MX). The present invention breaks through the limitations of traditional blood pressure monitoring methods, does not require compression of arteries, and provides a non-invasive, continuous blood pressure monitoring method, which is particularly suitable for daily wear and long-term monitoring. In addition, the operation method of the present invention is simple, the materials are easily available and low-cost, the biocompatibility is strong, the wearing is comfortable, the robustness is high, and the overall device is lightweight, similar to a Band-Aid. It has great innovation and advantages, and shows great application prospects in wearable pulse wave detection.

[0077] The specific embodiments of the present invention are further described below.

[0078] See also Figure 1, a method for manufacturing a pulse wave sensor, comprising the following steps:

[0079] 1) Wash the cut sponge with acetone and dry it at room temperature for 1 hour;

[0080] 2) After adding 10 mM Tris-HCl buffer to dopamine hydrochloride, soak the sponge cleaned in step 1) in the solution;

[0081] 3) taking out the sponge (denoted as PDA@SPG) fully soaked in the PDA solution in step 2) and drying it in an electric constant temperature incubator;

[0082] 4) Weigh the multi-walled carbon nanotubes and pour them into a small glass bottle, then drop isopropyl alcohol (AR, ≥99.7%) into the bottle, shake it thoroughly and place the bottle in an ultrasonic cleaner to make the CNTs fully dispersed in the isopropyl alcohol;

[0083] 5) placing the PDA@SPG dried in step 3) into the CNT dispersion prepared in step 4) and soaking it;

[0084] 6) taking the sponge soaked in the CNT dispersion in step 5) out of the glass bottle, placing it in an electric constant temperature incubator to dry it to obtain a conductive sponge (denoted as PDA-CNT@SPG);

[0085] 7) Use polyimide copper film to form interdigital electrodes and then plate them with gold to serve as a conductive circuit;

[0086] 8) Plasma-bonding the upper and lower PDMS sheets, aligning the conductive sponge obtained in step 6) with the conductive circuit obtained in step 7), and encapsulating them to obtain a pulse wave sensor;

[0087] In step 1), the sample should be dried at room temperature for about 1 hour;

[0088] In step 2), the Tris-HCl buffer is alkaline and the immersion time should be greater than 1 h;

[0089] In step 3), the soaked sponge is dried at a temperature between 50 and 100°C and the drying time is about 1 hour;

[0090] In step 4), CNTs should be dispersed in isopropanol, and the mass fraction of CNTs in the dispersion is about 0.1wt%-10wt%;

[0091] In step 5), the soaking time of the treated sample in the dispersion shall not be less than 10 minutes;

[0092] In step 6), the sample drying temperature is between 40°C and 100°C, and the drying time should be greater than 1 minute;

[0093] In step 7), the width of the interdigitated electrode channel should be less than 1 cm, and the width of a single electrode should be greater than 20 microns;

[0094] In step 8), at least one side of the upper and lower PDMS films needs to be plasma treated, and then the two layers are pressurized and bonded together. The greater the pressure, the tighter the bonding.

[0095] The embodiment of the present invention uses melamine sponge, polydopamine (PDA) and carbon nanotubes (CNT) as the main materials, obtains an ultra-light conductive sponge with a mass of only 3.3 mg by simple soaking and drying, and uses two layers of transparent PDMS film to encapsulate it to obtain a flexible resistive pressure sensor with a thickness of only 290 μm. The sensor has good linear response (sensitivity 6.23 kPa-1) and ultra-low detection limit (less than 100 Pa) in a wide pressure range (0-350 kPa), and the electrical response is still stable after 1500 cycles of compression, and has a very short response time (100 ms).

[0096] Blood pressure monitoring method based on pulse wave signal

[0097] The obtained pulse wave sensor is used to collect two segments of human pulse wave signals. The signal is collected through the pulse wave sensor. The cuff is thin and breathable, has no redundant wires and tubes, can be worn for a long time, and realizes real-time blood pressure monitoring.

[0098] Placing the sensor at the carotid artery and radial artery can obtain high-quality pulse wave signals. After data processing, it can display the main wave peak of each beat and calculate the heart rate. The sensor can stably measure pulse waves for a long time, and its performance is still stable after 30 days of placement, showing that it has great application prospects in wearable pulse wave detection. The back-end processing system includes an acquisition module and a data processing module, which are used to collect human pulse waves and extract and calculate key parameters such as pulse wave transmission time (PTT), pulse wave transmission velocity (PWV) and pulse wave main peak value (MX); the signal processing module calculates the first-level blood pressure value according to the blood pressure parameter model BP=aPWV^2+bPWV+cln(PWV)+d, where a, b, c, d are model parameters, calculated by nonlinear regression of experimental data. Then, the first-level blood pressure value and the measured MX are imported into the trained extreme learning machine (ELM) blood pressure model, MX is normalized, and the normalized pulse wave main peak value and other physiological parameters (such as heart rate, weight, etc.) are combined into the input matrix H. Set the number of hidden layer nodes of ELM to 1, randomly generate hidden layer weights W and hidden layer biases B, and calculate the hidden layer output matrix G. The formula is expressed as H = σ(WX+B), where σ is the activation function. Once the output H of the hidden layer is determined, ELM solves the output weight β by the least squares method, so that the predicted value Y' of the output layer is close to the true target value Y. The formula is: β = (H T H+λI) -1 H T Y where λ is the regularization parameter, I is the identity matrix, and Y is the target output value. For the experimental sample data X, ELM uses the same hidden layer weights and biases, as well as the trained output weights β for prediction. The formula is expressed as Y'=σ(WX+B)β. The robustness of ELM enables it to adapt to the inaccuracy of the data by adjusting the output weights β, thereby making effective predictions. This is because the core advantage of ELM is that it can quickly solve the output weights through analytical methods rather than through an iterative training process, which reduces the dependence on the quality of the input data. In addition, the generalization ability of ELM also helps to maintain the stability and accuracy of the prediction in the face of inaccurate inputs. This method has further improved the calculation accuracy compared to the original PWV linear model, quadratic model, logarithmic model, etc.; in the precise measurement of blood pressure, the learning method of the extreme learning machine is adopted, which has greatly improved the calculation accuracy.

[0099] Perform data preprocessing on the collected pulse wave signal to obtain PTT, PWV and MX;

[0100] According to the blood pressure parameter model, 100 sets of training data were substituted, including the real blood pressure values ​​and the measured PWV. Origin was used to perform nonlinear regression calculations on the data to obtain the optimal values ​​of the model parameters a, b, c and d, and a was -102.9, b was 19842, c was 1523 and d was -12;

[0101] The 100 sets of training data, the obtained first-level blood pressure prediction values ​​and MX were imported into ELM for training. The number of hidden layer nodes of ELM was set to 1, and the hidden layer weights W and hidden layer biases B were randomly generated.

[0102] Calculate the hidden layer output matrix G and determine the output H of the hidden layer. ELM solves the output weight β by the least squares method.

[0103] The obtained first-level blood pressure value and the obtained MX are used as input feature parameters and imported into the trained ELM blood pressure model to obtain the final blood pressure prediction value, which is expressed by the formula Y'=σ(WX+B)β.

[0104] Example 1

[0105] The method for making a pulse wave sensor comprises the following steps:

[0106] 1. Wash the cut sponge with acetone and dry it at room temperature for 1 hour;

[0107] 2. After adding dopamine hydrochloride to 10 mM Tris-HCl buffer at pH 8.5, soak the cleaned sponge in the solution for 24 h;

[0108] 3. Take out the sponge (denoted as PDA@SPG) fully soaked in 2 mg / mL PDA solution and dry it in an electric constant temperature incubator (Shanghai Yiheng Scientific Instrument Co., Ltd., model DHP-9052) at 70°C for 4 h;

[0109] 4. Weigh 50 mg of multi-walled carbon nanotubes and pour them into a small glass bottle. Then, drop 5 g of isopropanol (AR, ≥ 99.7%) into the bottle. After shaking thoroughly, place the bottle in an ultrasonic cleaner (Skymen, model JP-0205) for 30 min to fully disperse the CNTs in the isopropanol. The mass fraction of CNTs in the dispersion is about 1 wt%;

[0110] 5. Place the dried PDA@SPG into 1wt% CNT dispersion (dispersant is isopropanol) and soak for 10 hours;

[0111] 6. Take the sponge soaked in the CNT dispersion out of the glass bottle, put it into an electric constant temperature incubator at 70°C and dry it for 2 hours to obtain a conductive sponge (denoted as PDA-CNT@SPG);

[0112] 7. Use polyimide copper film to form interdigital electrodes and then plate them with gold to serve as a conductive circuit;

[0113] 8. Use the upper and lower PDMS sheets to bond after plasma treatment, align the obtained conductive sponge with the conductive circuit and then encapsulate to obtain a pulse wave sensor;

[0114] The blood pressure monitoring method based on pulse wave signal comprises the following steps:

[0115] Using the obtained pulse wave sensor to collect two segments of human body pulse wave signals;

[0116] Perform data preprocessing on the collected pulse wave signal to obtain PTT, PWV and MX;

[0117] The first-level blood pressure value is calculated according to the blood pressure parameter model;

[0118] The obtained first-level blood pressure value and the obtained MX are used as input feature parameters and imported into the trained ELM blood pressure model to obtain the final blood pressure prediction value.

[0119] This embodiment uses melamine sponge, polydopamine (PDA) and carbon nanotubes (CNT) as the main materials, and obtains an ultra-light conductive sponge with a mass of 3.3 mg by simple immersion and drying. It is encapsulated with two layers of transparent PDMS film to obtain a flexible resistive pressure sensor with a thickness of only 290 μm. After the pulse wave signal is collected by the sensor, the acquisition module obtains the pulse wave signal and extracts the pulse wave transit time (PTT), pulse wave velocity (PWV) and pulse wave main peak value (MX). The signal processing module uses the multi-modulus blood pressure parameter model to calculate the first-level blood pressure value, and then combines it with MX and inputs it into the trained ELM blood pressure model to obtain the final blood pressure prediction value.

[0120] Table 1 is the experimental data of systolic blood pressure of 10 test volunteers using the embodiment of the present invention, and Table 2 is the experimental data of diastolic blood pressure of 10 test volunteers using the embodiment of the present invention.

[0121] Table 1

[0122]

[0123]

[0124] Table 2

[0125]

[0126] The above content is a further detailed description of the present invention in combination with specific / preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, without departing from the concept of the present invention, it can also make several substitutions or modifications to these described embodiments, and these substitutions or modifications should be regarded as belonging to the protection scope of the present invention. In the description of this specification, the description of the reference terms "an embodiment", "some embodiments", "preferred embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in combination with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily target the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In the absence of mutual contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples. Although the embodiments of the present invention and their advantages have been described in detail, it should be understood that various changes, substitutions and modifications can be made herein without departing from the scope of protection of the patent application.

Claims

1. A pulse wave sensor, characterized in that: include: Melamine sponge; A polydopamine (PDA) layer, covering the melamine sponge, for forming a conductive layer; Carbon nanotubes (CNTs), dispersed in the polydopamine (PDA) layer to enhance conductivity and sensitivity; A conductive circuit, comprising a polyimide copper film and interdigital electrodes formed on the polyimide copper film; wherein the interdigital electrodes cooperate with the polydopamine (PDA) layer and carbon nanotubes (CNT) to transmit electrical signals; The upper and lower PDMS films encapsulate the melamine sponge, polydopamine (PDA) layer, carbon nanotubes (CNT) and conductive circuits to provide flexibility and protection.

2. The pulse wave sensor according to claim 1, characterized in that: The PDMS film was plasma treated to enhance adhesion.

3. The pulse wave sensor according to claim 1 or 2, characterized in that: The width of the interdigitated electrode channel should be less than 1 cm, and the width of a single electrode should be greater than 20 microns.

4. A method for manufacturing a pulse wave sensor, characterized in that: The following steps are involved: S1. Prepare melamine sponge; S2, covering the melamine sponge with a polydopamine (PDA) layer to form a conductive layer; S3, dispersing carbon nanotubes (CNT) in the polydopamine (PDA) layer to enhance conductivity and sensitivity; S4, preparing a conductive circuit, including forming interdigital electrodes on the polyimide copper film; S5. Encapsulate the melamine sponge, polydopamine (PDA) layer, carbon nanotube (CNT) and conductive circuit between upper and lower PDMS films to provide flexibility and protection; wherein the interdigitated electrodes cooperate with the polydopamine (PDA) layer and carbon nanotube (CNT) to transmit electrical signals.

5. The method according to claim 4, characterized in that: In step S2, the formation of the polydopamine (PDA) layer comprises adding Tris-HCl buffer to dopamine hydrochloride, and then soaking the sponge cleaned in step S1 in the solution for a soaking time of more than 1 hour.

6. The method according to claim 4 or 5, characterized in that: In step S3, multi-walled carbon nanotubes (CNTs) are poured into a container, isopropanol is added and ultrasonic treatment is performed to fully disperse the CNTs in the isopropanol, the mass fraction of CNTs in the dispersion is 0.1wt% to 10wt%, and the melamine sponge covered with a polydopamine (PDA) layer is immersed in the CNT dispersion and then dried.

7. The method according to claim 6, characterized in that: The immersion time in the dispersion is not less than 10 minutes, the drying temperature is 40°C-100°C, and the drying time is not less than 1 minute.

8. The method according to any one of claims 4 to 7, characterized in that: In step S5, at least one side of the upper and lower PDMS films is treated with plasma, and then the two PDMS films are pressure-bonded.

9. A blood pressure monitoring device based on a pulse wave signal, characterized in that: The invention comprises a pulse wave sensor and a data processing system as claimed in any one of claims 1 to 3.

10. A blood pressure monitoring method based on pulse wave transit time, characterized in that The following steps are involved: S1. Collecting a pulse wave signal of a human body using the pulse wave sensor as described in any one of claims 1 to 3; S2, preprocessing the collected pulse wave signal to extract the pulse wave transit time (PTT), pulse wave velocity (PWV) and pulse wave main peak value (MX); S3, using a multimodal blood pressure parameter model to calculate the first-level blood pressure value, the model is BP = aPWV^2 + bPWV + cln(PWV) + d, where a, b, c, d are model parameters calculated by nonlinear regression of experimental data; S4, taking the first-level blood pressure value obtained in step S3 and the MX obtained in step S2 as input feature parameters, and importing them into the trained extreme learning machine (ELM) blood pressure model; S5, normalizing MX in the ELM model, and combining the normalized MX with other physiological parameters into an input matrix H; S6. Set the number of hidden layer nodes of ELM to 1, randomly generate hidden layer weights W and hidden layer biases B, and calculate the hidden layer output matrix G, which is expressed as H = σ(WX+B), where σ is the activation function; S7. Once the output H of the hidden layer is determined, ELM uses the least squares method to solve the output weight β, so that the predicted value Y' of the output layer is close to the true target value Y, and the formula is expressed as β = (H T H+λI) -1 H T Y, where λ is the regularization parameter and I is the identity matrix; S8. Use the output weight β obtained in S7 to predict the experimental sample data X, and the formula is expressed as Y'=σ(WX+B)β.

Images