Single-point radial artery wave-based cuffless blood pressure measurement apparatus and wearable device

Abstract

The present application relates to a single-point radial artery wave-based cuffless blood pressure measurement apparatus and a wearable device. The apparatus comprises: a pulse sensor, an analog-to-digital conversion unit, a micro-processing control unit, and a terminal device. The pulse sensor acquires a single-point pulse wave signal of a human radial artery; the analog-to-digital conversion unit performs analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal; the micro-processing control unit receives the single-point pulse wave digital signal for processing, and analyzes the single-point pulse wave digital signal by means of a cuffless blood pressure measurement model to resolve a human blood pressure value; and the terminal device is used for performing AI analysis, data display, waveform display, health report output and / or early warning and prompting on continuous dynamic changes in the single-point pulse wave digital signal and a blood pressure value. The present application achieves the continuous measurement of pulse waves, human pulse monitoring and non-invasive cuffless continuous blood pressure monitoring, effectively improves the early diagnosis of hypertensive diseases, and reduces mortality from hypertensive diseases.

Description

Cuffless blood pressure detection device and wearable device based on single-point radial artery wave

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application (application number 202310725128.3) filed on June 19, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present application relates to a cuffless blood pressure detection device and a wearable device based on a single-point radial artery wave, and relates to the field of medical detection technology. Background Art

[0004] Among the many physiological indicators associated with cardiovascular health, blood pressure plays a crucial role in the diagnosis of cardiovascular disease. Active and continuous blood pressure monitoring can effectively prevent cardiovascular-related deaths. However, current global management of hypertension remains plagued by low diagnosis rates, high morbidity, and poor control rates. Continuous blood pressure monitoring is a cost-effective and effective means of improving early diagnosis and control rates of hypertension, and reducing mortality rates. However, due to limitations in detection principles and instrument configuration, traditional non-invasive blood pressure monitoring methods require the use of an inflatable cuff and a fixed posture for blood pressure measurement. Consequently, they can only measure blood pressure intermittently and are unable to record continuous, dynamic blood pressure changes during daily activities and sleep. Furthermore, multiple daily blood pressure measurements are a significant burden for the tester, leading to low user compliance, insufficient daily blood pressure measurement, and inadequate disease management.

[0005] In order to achieve portable, comfortable and long-term blood pressure monitoring, researchers have proposed cuffless blood pressure measurement technology and equipment. Common non-invasive continuous blood pressure detection methods in the existing technology include arterial tension method, volume compensation method, pulse wave transit time (PTT) method and pulse wave characteristic method. Among them, the pulse wave transit time method and pulse wave characteristic parameter method are relatively convenient to measure, highly comfortable and have more ideal results. However, the existing continuous blood pressure monitoring model still has many shortcomings. The stability and accuracy of clinical monitoring are not ideal, and it cannot overcome the influence of individual differences and complex factors during long-term monitoring. It has not yet achieved the level of widespread clinical acceptance.

[0006] In summary, although cuffless continuous blood pressure monitoring technology and equipment have made great progress, the following problems still exist: (1) Most devices collect photoplethysmography (PPG) signals, but PPG signals can only indirectly reflect pulse wave changes, and there are morphological differences between PPG signals and pulse wave signals; (2) Most devices require multi-point measurements, such as combining PPG signals and electrocardiogram (ECG) signals to calculate PTT and then estimate blood pressure, but the measurement of ECG signals requires electrodes or contact between hands to form a closed-loop path, making the device less user-friendly; (3) Most algorithms only use one or several feature information of physiological signals such as pulse waves as input to the blood pressure prediction model. These feature information are manually designed and rarely or not consider the overall information of the original pulse wave. The original pulse waveform contains valuable information describing cardiovascular function, and the manually designed features may be omitted; (4) Most devices are not integrated into a wearable style, which is cumbersome to use.

[0007] Invention Disclosure

[0008] The present application aims to solve at least one of the technical problems existing in the prior art. To this end, in response to the above-mentioned problem, the purpose of the present application is to provide a cuffless blood pressure detection device and wearable device based on a single-point radial artery wave that can realize cuffless human blood pressure detection and long-term blood pressure tracking and monitoring.

[0009] In order to achieve the above-mentioned invention objectives, the technical solutions adopted in this application are:

[0010] In a first aspect, the present application provides a cuffless blood pressure detection device based on a single-point radial artery wave, the device comprising a pulse sensor, an analog-to-digital conversion unit, and a microprocessor control unit;

[0011] The pulse sensor is configured to collect a single-point pulse wave signal from the radial artery of the human body;

[0012] The analog-to-digital conversion unit is configured to perform analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal;

[0013] The microprocessor control unit is configured to receive the single-point pulse wave digital signal for processing, and analyze the processed single-point pulse wave digital signal through a cuffless blood pressure detection model to resolve the human blood pressure value.

[0014] Furthermore, the pulse sensor adopts a flexible pressure sensor based on a multi-level branch structure, a resistive pulse sensor, a capacitive pulse sensor, a piezoelectric pulse sensor or a triboelectric pulse sensor.

[0015] Furthermore, the training of the cuffless blood pressure detection model includes:

[0016] Obtaining training data, including single-point pulse wave signals, pulse wave characteristics, human physiological and pathological characteristics, and reference blood pressure data;

[0017] Establishing a cuffless blood pressure detection model, wherein the cuffless blood pressure detection model includes a deep learning model and a theoretical formula model;

[0018] The pulse wave signal, pulse wave characteristics, human physiological characteristics and reference blood pressure data are used as inputs of a cuffless blood pressure detection model to train the cuffless blood pressure detection model, wherein the pulse wave signal, human physiological characteristics and reference blood pressure data are used as input features of a deep learning model, and the pulse wave characteristics are used as input features of a theoretical formula model.

[0019] Furthermore, the pulse wave characteristics are obtained by extracting a single-point pulse wave signal, including waveform characteristics and pulse wave characteristics related to the cardiovascular health of the human body, wherein the waveform characteristics include peak height, peak width, peak position, and the slope of the peak rising or falling edge; the pulse wave characteristics related to the cardiovascular health of the human body include reflected wave conduction time RWTT, radial artery augmentation index AI r and heart rate HR.

[0020] Furthermore, the backbone network of the deep learning model adopts a time series prediction model based on Transformer or other neural network architecture, including TimesNet, ETSformer, FEDformer, Pyraformer, Autoformer, Informer or Reformer.

[0021] Furthermore, the human physiological characteristics include the patient's age, gender, height and weight.

[0022] Furthermore, the theoretical formula model is established based on the MK equation with the reflected wave transmission time RWTT as the core, and the formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d

[0023] Where BP is the human blood pressure value, and a, b, c, and d are model parameters, which are calculated by the deep learning model based on the single-point pulse wave signal, human physiological characteristics, and reference blood pressure values.

[0024] Furthermore, it also includes a terminal device and a communication unit, wherein the communication unit adopts a wired or wireless transmission method to enable the microprocessor control unit to transmit data with the terminal device; wherein,

[0025] The terminal device is also provided with the cuffless blood pressure detection model, which is used to analyze the single-point pulse wave digital signal to parse out the human blood pressure value, and / or

[0026] The terminal device is used to perform AI analysis, data display, waveform display and / or health report output on the continuous dynamic changes of single-point pulse wave digital signals and blood pressure values, and can provide text, sound and other prompts and warnings for abnormal situations.

[0027] In a second aspect, the present application provides a wearable device, comprising a wristband structure and a cuffless blood pressure detection device based on a single-point radial artery wave as described in any one of the first aspects of the present application, wherein the cuffless blood pressure detection device is arranged in the wristband structure.

[0028] Furthermore, the display screen is provided on the top of the card box body, the display screen is connected to the microprocessor control unit, the cuffless blood pressure detection device is provided in the card box body, the card box base is detachably connected to the card box body through the rotating shaft, the card box base is fixedly provided on the wristband, and the card box base is provided with a detection hole, and the pulse sensor detects the single-point pulse wave signal of the human radial artery through the detection hole.

[0029] In a third aspect, the present application provides a blood pressure monitoring method, comprising the following steps:

[0030] 1) Collecting single-point pulse wave signals from the human radial artery;

[0031] 2) performing analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal;

[0032] 3) receiving the single-point pulse wave digital signal for processing, and analyzing the processed single-point pulse wave digital signal through a cuffless blood pressure detection model to parse out the human blood pressure value.

[0033] Furthermore, the training of the cuffless blood pressure detection model includes:

[0034] Obtaining training data, including single-point pulse wave signals, pulse wave characteristics, human physiological and pathological characteristics, and reference blood pressure data;

[0035] Establishing a cuffless blood pressure detection model, wherein the cuffless blood pressure detection model includes a deep learning model and a theoretical formula model;

[0036] The pulse wave signal, pulse wave characteristics, human physiological characteristics and reference blood pressure data are used as inputs of a cuffless blood pressure detection model to train the cuffless blood pressure detection model, wherein the pulse wave signal, human physiological characteristics and reference blood pressure data are used as input features of a deep learning model, and the pulse wave characteristics are used as input features of a theoretical formula model.

[0037] Furthermore, the pulse wave characteristics are obtained by extracting a single-point pulse wave signal, including waveform characteristics and pulse wave characteristics related to the cardiovascular health of the human body, wherein the waveform characteristics include peak height, peak width, peak position, and the slope of the peak rising or falling edge; the pulse wave characteristics related to the cardiovascular health of the human body include reflected wave conduction time RWTT, radial artery augmentation index AI r and heart rate HR.

[0038] Furthermore, the backbone network of the deep learning model adopts a time series prediction model based on Transformer or other neural network architecture, including TimesNet, ETSformer, FEDformer, Pyraformer, Autoformer, Informer or Reformer.

[0039] Furthermore, the theoretical formula model is established based on the MK equation with the reflected wave transmission time RWTT as the core, and the formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d

[0040] Where BP is the human blood pressure value, and a, b, c, and d are model parameters, which are calculated by the deep learning model based on the single-point pulse wave signal, human physiological characteristics, and reference blood pressure values.

[0041] Furthermore, the human physiological characteristics include the patient's age, gender, height and weight.

[0042] Furthermore, the blood pressure monitoring method is used in non-invasive body measurement and long-term health tracking monitoring of cardiovascular and cerebrovascular diseases, sleep apnea syndrome, vascular plaque and sclerosis, stroke, hypertension, and obesity. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present application. Throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0044] FIG1 is a schematic structural diagram of a cuffless blood pressure detection device according to an embodiment of the present application;

[0045] FIG2 is a schematic structural diagram of a flexible sensor according to an embodiment of the present application;

[0046] FIG3 is a flowchart of the use of the cuffless blood pressure detection device according to an embodiment of the present application;

[0047] FIG4 is a schematic diagram of a cuffless blood pressure detection model according to an embodiment of the present application;

[0048] FIG5 is a schematic diagram of radial artery pressure waves and related characteristic points according to an embodiment of the present application;

[0049] FIG6 is a schematic diagram of wearing a wearable device for a radial artery wave acquisition experiment according to an embodiment of the present application;

[0050] Figures 7(a) and (b) are schematic diagrams of the mechanical structure of the wearable device according to an embodiment of the present application.

[0051] Best Mode for Carrying Out the Invention

[0052] It should be understood that the terms used herein are for the purpose of describing specific example embodiments only and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "comprise", "include", "contain" and "have" are inclusive and therefore specify the presence of stated features, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be performed in the specific order described or illustrated, unless the order of execution is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0053] Although the terms first, second, third, etc. can be used in the text to describe multiple elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms can only be used to distinguish an element, component, region, layer or section from another region, layer or section. Unless the context clearly indicates otherwise, terms such as "first", "second" and other numerical terms do not imply order or sequence when used in the text. Therefore, the first element, component, region, layer or section discussed below can be referred to as the second element, component, region, layer or section without departing from the teaching of the example embodiments.

[0054] For ease of description, spatially relative terms may be used herein to describe the relationship of one element or feature relative to another element or feature as shown in the figures, such as "inside," "outside," "inner side," "outer side," "lower," "upper," etc. Such spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures.

[0055] Since existing blood pressure detection devices are less user-friendly, the blood pressure detection input features are not comprehensive, and the use is relatively cumbersome. The present application provides a cuffless blood pressure detection device and wearable device based on a single-point radial artery wave, the device includes a pulse sensor, an analog-to-digital conversion unit, a microprocessor control unit, and a terminal device; the pulse sensor is configured to collect a single-point pulse wave signal from the human radial artery; the analog-to-digital conversion unit is configured to perform analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal; the microprocessor control unit is configured to receive the single-point pulse wave digital signal for processing, and analyze the processed single-point pulse wave digital signal through a cuffless blood pressure detection model to parse out the human blood pressure value; the terminal device is used to perform AI analysis, data display, waveform display, and health report output on the continuous dynamic changes of the single-point pulse wave digital signal and the blood pressure value, and to provide text, sound, and other prompts and warnings for abnormal situations. Therefore, the present application can monitor the multimodal information of cardiovascular health related to the user's pulse wave, with high adaptability and accuracy.

[0056] The following describes exemplary embodiments of the present application in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0057] Embodiment 1: As shown in FIG1 , the cuffless blood pressure detection device based on a single-point radial artery wave provided in this embodiment includes a pulse sensor 1, an analog-to-digital conversion unit 2, a microprocessor control unit 3, a communication unit 4, and a terminal device 5;

[0058] Pulse sensor 1, used to collect a single-point pulse wave signal from the human radial artery, wherein the single-point pulse wave signal is a pulse wave signal at a single position on the radial artery;

[0059] Analog-to-digital conversion unit 2, used to perform analog-to-digital conversion on the single-point pulse wave signal to obtain a digital signal of the single-point pulse wave;

[0060] The microprocessor control unit 3 is used to receive the digital signal from the analog-to-digital conversion unit 2 for processing, and input the pulse wave digital signal into a preset cuffless blood pressure detection model for analysis, parse out the human blood pressure value, and send the processing result to the communication unit 4 through the asynchronous serial interface.

[0061] The communication unit 4 enables information communication between the microprocessor control unit 3 and the terminal device 5 through wired or wireless communication.

[0062] In a preferred embodiment, the pulse sensor 1 may adopt a flexible pressure sensor based on a multi-level branch structure, or may adopt other commercial pulse sensors based on resistance, capacitance, piezoelectric or triboelectric types, etc., which are not limited here.

[0063] Furthermore, as shown in FIG2 , the multi-level branched structure pressure sensor provided in this embodiment includes a flexible substrate 11, a first sensor signal detection electrode 12, a second sensor signal electrode 13, a primary structure layer 14, and a secondary structure layer 15. The primary structure layer 14 and the secondary structure layer 15 are stacked on top of the flexible substrate 11, and the first sensor signal detection electrode 12 and the second sensor signal electrode 13 are spaced apart between the secondary structure layer 15 and the flexible substrate 11. The surface of the secondary structure layer 15 is coated with an active layer material for pressure sensing. The active layer material can be reduced graphene oxide material, carbon nanotube material, or the like. Microstructures are disposed beneath both the primary structure layer 14 and the secondary structure layer 15. The geometric shapes of the microstructures are, but are not limited to, hemispherical structures, pyramid structures, boss structures, or linear structures.

[0064] In a preferred embodiment, as shown in FIG3 , the cuffless blood pressure detection model is a systolic and diastolic pressure prediction model based on pulse wave data, wave features extracted from the pulse wave, and human physiological characteristics, and is based on the Moens-Korteweg (MK) equation with pulse wave transit time as the core and the Transformer architecture with pulse waveform as the core, with subject-aware capabilities. The training process is as follows:

[0065] S1. Collect single-point pulse wave signals and reference blood pressure values ​​from the human radial artery.

[0066] In this embodiment, the blood pressure of a human body is measured as a reference blood pressure value, wherein the reference blood pressure value is the diastolic pressure and the systolic pressure measured using an existing electronic sphygmomanometer and / or a mercury sphygmomanometer.

[0067] S2. Perform data preprocessing on the single-point pulse wave signal.

[0068] In this embodiment, data preprocessing includes performing operations such as removing high-frequency noise, removing low-frequency baseline drift, signal alignment, and signal segmentation on the single-point pulse wave signal.

[0069] Furthermore, this embodiment uses wavelet transform to filter out high-frequency noise and low-frequency baseline drift in the original pulse wave signal. Specifically, the baseline drift frequency band and the high-frequency noise frequency band are suppressed, and the wavelet coefficients in other frequency bands are used to reconstruct the pulse wave signal to obtain a filtered and denoised pulse wave signal. By identifying the starting point of a pulse wave cycle, the pulse wave cycle is segmented and the signal is aligned.

[0070] S3. Perform feature point recognition on the pulse wave signal after data preprocessing to extract wave features in the pulse wave signal.

[0071] In this embodiment, the wave features in the pulse wave signal include waveform features and pulse wave features related to the cardiovascular health of the human body.

[0072] Furthermore, the waveform characteristics include peak height, peak width, peak position, slope of peak rising / falling edge, etc.

[0073] Furthermore, pulse wave characteristics related to cardiovascular health include reverse pulse transit time (RWTT), radial artery augmentation index (AI), r ) and heart rate (HR), etc.

[0074] S4. Obtain training data based on the pulse wave signal, pulse wave characteristics, human physiological and pathological characteristics, and reference blood pressure data.

[0075] In this embodiment, human physiological characteristics include the patient's age, gender, height and weight, etc., which are used together with the pulse wave signal as input features of the deep learning model, and the pulse wave features are used as input features of the theoretical formula model.

[0076] S5. Establish a cuffless blood pressure detection model

[0077] In this embodiment, the cuffless blood pressure detection model includes two parts: the first part is an artificial intelligence model or a deep learning model, and the second part is a theoretical formula model.

[0078] The backbone network of the artificial intelligence model can adopt any time series prediction model based on Transformer or other neural network architecture, including TimesNet, ETSformer, FEDformer, Pyraformer, Autoformer, Informer, Reformer, etc. This embodiment does not limit the specific model used by the above backbone network.

[0079] As shown in Figure 4, an artificial intelligence model is constructed using a time series prediction model based on the Transformer architecture, and a theoretical formula model is constructed based on the MK equation. The artificial intelligence model uses pulse wave signals and human physiological characteristics to calculate the parameters in the theoretical formula model, and then uses the pulse wave characteristics through the theoretical formula model to calculate systolic and diastolic blood pressures, specifically:

[0080] Figure 5 shows the pulse wave feature points that need to be identified in single pulse wave cuffless blood pressure testing in this embodiment. This application combines wavelet transform and differential method to identify and calculate feature points of the preprocessed pulse waveform, including: the pulse wave starting point B, the peak P1 of the main pulse wave and the corresponding time t1 and amplitude R1, the peak P2 of the secondary pulse wave and the corresponding time t2 and amplitude R2, the peak P3 of the third pulse wave and the corresponding time t3 and amplitude R3, and the pulse wave period T.

[0081] These characteristic points are used to calculate pulse wave characteristics, including: reflected wave conduction time RWTT, radial artery enhancement index AI r And heart rate HR, etc. Among them, RWTT is the time difference between t2 and t1 (RWTT = t2-t1), AI r is the ratio between R2 and R1 (AI r =R2 / R1), HR is the inverse of the pulse wave period T (HR=1 / T).

[0082] According to the pulse wave characteristics and the corresponding blood pressure values, based on the MK equation, a theoretical formula model with the reflected wave transmission time RWTT as the core is established. The formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d

[0083] The theoretical formula model is based on pulse wave characteristics (RWTT, AI r , HR) to predict systolic blood pressure (SBP) and diastolic blood pressure (DBP). Here, blood pressure is uniformly represented by BP. a, b, c, and d are parameters in the theoretical formula model, calculated using a deep learning model based on actual human pulse wave signals and physiological characteristics.

[0084] S6. Divide the obtained training data into training set, validation set and test set according to a certain ratio. The training set and validation set are used to train and optimize the hyperparameters of the cuffless blood pressure detection model, and the test set is used to test the generalization ability of the model obtained after training and hyperparameter tuning.

[0085] In this embodiment, a data set is constructed based on the pulse wave signal, the pulse wave features extracted from the pulse wave signal, the physiological characteristics of the human body, and the reference blood pressure value, and is used as the input of the cuffless blood pressure detection model to train the cuffless blood pressure detection model. The output of the model is the human blood pressure (including systolic pressure and diastolic pressure).

[0086] Furthermore, the data set of this embodiment is divided into a training set, a validation set, and a test set in a ratio of 8:1:1. The training set and the validation set are used to train and select parameters for the established cuffless blood pressure detection model, and the test set is used to test the generalization ability of the model. The data segmentation adopts the method of inter-individual segmentation to ensure that each set of data does not contain data from the same subject, thereby avoiding information leakage. During the training of the model, the Adam optimizer is used to update the parameters with a learning rate of 0.0002. On the training set, by calculating the error between the estimated blood pressure value and the true blood pressure value, the model parameters are updated using the back propagation algorithm, so that the estimated blood pressure value becomes closer and closer to the true blood pressure value.

[0087] Furthermore, since blood pressure prediction is a regression problem, generally only the encoder part of the time series prediction model based on the Transformer architecture is needed, and the decoder part is deleted. The deep learning model combines the deep waveform (deep waveform features refer to the high-level semantic information extracted from the pulse wave signal by the artificial intelligence algorithm. These features represent the human cardiovascular status information and are helpful for blood pressure prediction. They are learned by the algorithm from the data) features of the encoded pulse wave time series data with the blood pressure values ​​corresponding to the human physiological characteristics. The fusion method is to use a multi-layer fully connected neural network, which can better weight the analysis of the proportion of different information. In each blood pressure test, the constant parameters of the theoretical formula model calculated by the deep learning model may also be different according to the different pulse waves of the current user, thereby realizing adaptive calculation and personalized monitoring of blood pressure.

[0088] In a preferred embodiment, the terminal device 5 can be a mobile phone, a computer, an LCD screen, a cloud server or other wired / wireless device. The terminal device 5 can display pulse waves and blood pressure values. The terminal device 5 can be used to implement functions such as AI analysis, data display, waveform display and health report sharing. It can be used to visualize the information on the cardiovascular health status of the human body, and realize human health monitoring by performing AI analysis on the continuous dynamic changes of pulse waves and blood pressure values. It can also provide text, sound and other prompts and warnings for abnormal situations.

[0089] Furthermore, the terminal device 5 may also be preset with or without a cuff blood pressure detection model to calculate the diastolic and systolic blood pressures and display the test results at the same time.

[0090] In a preferred embodiment, the communication unit 4 can use wired and wireless transmission methods such as Bluetooth or ZigBee to send and receive data.

[0091] Embodiment 2: As shown in FIG6 , this embodiment further provides a wearable device 6 including the cuffless blood pressure detection device based on the single-point radial artery wave of embodiment 1.

[0092] As shown in FIG7 , the wearable device 6 may adopt a card box type wristband structure, including a card box 61 and a wristband 62 , and the card box 61 and the wristband 62 are detachably arranged.

[0093] In this embodiment, the card cartridge 61 includes a card cartridge body 611, a display screen 612, a card cartridge base 613, and a detachable shaft 614. A cuffless blood pressure monitoring device based on a single-point radial artery wave is housed within the card cartridge body 611. A display screen 612 is located on top of the card cartridge body 611, which is connected to the microprocessor control unit 3 for displaying blood pressure results. The card cartridge base 613 is connected to the card cartridge body 611 via a detachable shaft 614. The card cartridge base 613 is fixed to the wristband 62, and the detachable shaft 614 enables the card cartridge body 611 and the card cartridge base 613 to be removed and rotated. The card cartridge base 613 is provided with a detection hole 615, which provides a window for the user to locate the radial artery. The pulse sensor 1 is located at the detection hole 615.

[0094] This application also provides a blood pressure monitoring method, comprising the following steps:

[0095] 1) Collecting single-point pulse wave signals from the human radial artery;

[0096] 2) performing analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal;

[0097] 3) receiving the single-point pulse wave digital signal for processing, and analyzing the processed single-point pulse wave digital signal through a cuffless blood pressure detection model to parse out the human blood pressure value.

[0098] Furthermore, the training of the cuffless blood pressure detection model includes:

[0099] Obtaining training data, including single-point pulse wave signals, pulse wave characteristics, human physiological and pathological characteristics, and reference blood pressure data;

[0100] Establishing a cuffless blood pressure detection model, wherein the cuffless blood pressure detection model includes a deep learning model and a theoretical formula model;

[0101] The pulse wave signal, pulse wave characteristics, human physiological characteristics and reference blood pressure data are used as inputs of a cuffless blood pressure detection model to train the cuffless blood pressure detection model, wherein the pulse wave signal, human physiological characteristics and reference blood pressure data are used as input features of a deep learning model, and the pulse wave characteristics are used as input features of a theoretical formula model.

[0102] Furthermore, the pulse wave characteristics are obtained by extracting a single-point pulse wave signal, including waveform characteristics and pulse wave characteristics related to the cardiovascular health of the human body, wherein the waveform characteristics include peak height, peak width, peak position, and the slope of the peak rising or falling edge; the pulse wave characteristics related to the cardiovascular health of the human body include reflected wave conduction time RWTT, radial artery augmentation index AI r and heart rate HR.

[0103] Furthermore, the backbone network of the deep learning model adopts a time series prediction model based on Transformer or other neural network architecture, including TimesNet, ETSformer, FEDformer, Pyraformer, Autoformer, Informer or Reformer.

[0104] Furthermore, the theoretical formula model is established based on the MK equation with the reflected wave transmission time RWTT as the core, and the formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d

[0105] Where BP is the human blood pressure value, and a, b, c, and d are model parameters, which are calculated by the deep learning model based on the single-point pulse wave signal, human physiological characteristics, and reference blood pressure values.

[0106] Furthermore, the human physiological characteristics include the patient's age, gender, height and weight.

[0107] The wearable device and cuffless blood pressure measurement method based on this application are widely suitable for non-invasive in-body measurement and long-term health tracking and monitoring of common cardiovascular and cerebrovascular diseases, sleep apnea syndrome, vascular plaque and sclerosis, stroke, hypertension, obesity, etc.

[0108] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In the description of this specification, the reference terms "a preferred embodiment", "further", "specifically", "in the present embodiment", etc. mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of this specification. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, 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, unless they are contradictory.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

[0110] Industrial Applications

[0111] Due to the adoption of the above technical solution, this application has the following characteristics:

[0112] 1. The cuffless blood pressure detection device of the present application can realize continuous detection of the pulse wave of the user's radial artery, and is suitable for continuously monitoring the pulse changes of the user during daily activities or sleep at night, realizing human pulse monitoring and non-invasive cuffless continuous blood pressure monitoring, effectively improving the early diagnosis of hypertension and reducing the mortality rate of hypertension.

[0113] 2. The cuffless blood pressure detection device of the present application uses a single-point pulse wave as input, and utilizes the single-point pulse wave data, the pulse wave features extracted from the pulse wave, and the physiological and pathological characteristics of the human body to achieve continuous blood pressure prediction. Compared with the existing technology, the present application can make full use of the multimodal information related to the user's cardiovascular health and has high adaptability and accuracy.

[0114] 3. Unlike the fixed dial of traditional wearable devices, the wearable device of the present application is not only portable, but the rotatable and detachable card box structure can provide a window for the user to determine the position of the radial artery, thereby ensuring that the collected pulse wave signal has high strength and accuracy.

[0115] 4. This application utilizes a flexible pulse sensor to detect the weak pulse wave signal at the radial artery, which has individual characteristic differences. By utilizing the radial artery pulse wave data, the waveform features extracted from the pulse wave, and the physiological characteristics of the human body, continuous blood pressure prediction with subject-awareness is achieved, thereby realizing human health monitoring and disease prediction and early warning.

[0116] In summary, this application can be widely used in blood pressure measurement and daily blood pressure monitoring.

Claims

1. A cuffless blood pressure detection device based on a single-point radial artery wave, wherein the device comprises a pulse sensor, an analog-to-digital conversion unit and a microprocessor control unit; The pulse sensor is configured to collect a single-point pulse wave signal of a human radial artery; The analog-to-digital conversion unit is configured to perform analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal; The microprocessor control unit is configured to receive the single-point pulse wave digital signal for processing, and analyze the processed single-point pulse wave digital signal through a cuffless blood pressure detection model to parse out the human blood pressure value.

2. According to claim 1, the cuffless blood pressure detection device based on single-point radial artery wave, wherein the pulse sensor adopts a flexible pressure sensor based on a multi-level branch structure, a resistive pulse sensor, a capacitive pulse sensor, a piezoelectric pulse sensor or a friction electric pulse sensor.

3. The cuffless blood pressure detection device based on single-point radial artery wave according to claim 1, wherein the training of the cuffless blood pressure detection model comprises: Obtaining training data, including single-point pulse wave signals, pulse wave characteristics, human physiological and pathological characteristics, and reference blood pressure data; Establishing a cuffless blood pressure detection model, wherein the cuffless blood pressure detection model includes a deep learning model and a theoretical formula model; The pulse wave signal, pulse wave features, human physiological characteristics and reference blood pressure data are used as inputs of a cuffless blood pressure detection model to train the cuffless blood pressure detection model, wherein the pulse wave signal, human physiological characteristics and reference blood pressure data are used as input features of a deep learning model, and the pulse wave features are used as input features of a theoretical formula model.

4. According to the cuffless blood pressure detection device based on single-point radial artery wave of claim 3, the pulse wave characteristics are obtained by extracting the single-point pulse wave signal, including waveform characteristics and pulse wave characteristics related to the cardiovascular health of the human body, wherein: Waveform characteristics include peak height, peak width, peak position, and the slope of the peak rising or falling edge; the pulse wave characteristics related to human cardiovascular health include reflected wave conduction time RWTT, radial artery augmentation index AI r and heart rate HR.

5. According to claim 3, the cuffless blood pressure detection device based on single-point radial artery wave, wherein the backbone network of the deep learning model adopts a time series prediction model based on Transformer or other neural network architecture, including TimesNet, ETSformer, FEDformer, Pyraformer, Autoformer, Informer or Reformer. 6 . The cuffless blood pressure detection device based on single-point radial artery wave according to claim 3 , wherein the human physiological characteristics include the patient's age, gender, height and weight.

7. The cuffless blood pressure detection device based on a single-point radial artery wave according to any one of claims 3 to 5, wherein the theoretical formula model is a theoretical formula model based on the MK equation with the reflected wave transmission time RWTT as the core, and the formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d In the formula, BP is the human blood pressure value, a, b, c, and d are model parameters, which are calculated by the deep learning model based on the single-point pulse wave signal, human physiological characteristics, and reference blood pressure values.

8. The cuffless blood pressure detection device based on a single-point radial artery wave according to any one of claims 1 to 3, further comprising a terminal device and a communication unit, wherein the communication unit adopts a wired or wireless transmission method to enable the microprocessor control unit to transmit data with the terminal device; wherein, The terminal device is also provided with the cuffless blood pressure detection model, which is used to analyze the single-point pulse wave digital signal to parse out the human blood pressure value, and / or The terminal device is used to perform AI analysis, data display, waveform display, health report output and / or early warning prompts on the continuous dynamic changes of single-point pulse wave digital signals and blood pressure values.

9. A wearable device, comprising a wristband structure and a cuffless blood pressure detection device based on a single-point radial artery wave according to any one of claims 1 to 5, wherein the cuffless blood pressure detection device is arranged in the wristband structure.

10. The wearable device according to claim 9, wherein the wristband structure comprises a card box and a wristband, the card box comprises a card box body, a display screen, a card box base and a rotating shaft, the display screen is arranged on the top of the card box body, the display screen is connected to the microprocessor control unit, the cuffless blood pressure detection device is arranged in the card box body, the card box base is detachably connected to the card box body through the rotating shaft, the card box base is fixedly arranged on the wristband, the card box base is provided with a detection hole, and the pulse sensor detects the single-point pulse wave signal of the radial artery of the human body through the detection hole.

11. The wearable device according to claim 9, wherein the human physiological characteristics include age, gender, height and weight of the patient.

12. The wearable device according to claim 9, wherein the theoretical formula model is a theoretical formula model based on the MK equation with the reflected wave transmission time RWTT as the core, and the formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d In the formula, BP is the human blood pressure value, a, b, c, and d are model parameters, which are calculated by the deep learning model based on the single-point pulse wave signal, human physiological characteristics, and reference blood pressure values.

13. The wearable device according to claim 9, further comprising a terminal device and a communication unit, wherein the communication unit adopts a wired or wireless transmission method to enable the microprocessor control unit to perform data transmission with the terminal device; wherein, The terminal device is also provided with the cuffless blood pressure detection model. The model is used to analyze the single-point pulse wave digital signal to parse the human blood pressure value, and / or The terminal device is used to perform AI analysis, data display, waveform display, health report output and / or early warning prompts on the continuous dynamic changes of single-point pulse wave digital signals and blood pressure values.

14. A blood pressure monitoring method comprising the following steps: 1) Collecting single-point pulse wave signals from the human radial artery; 2) performing analog-to-digital conversion on the single-point pulse wave signal to obtain a single-point pulse wave digital signal; 3) receiving the single-point pulse wave digital signal for processing, and analyzing the processed single-point pulse wave digital signal through a cuffless blood pressure detection model to parse out the human blood pressure value.

15. The blood pressure monitoring method according to claim 15, wherein the training of the cuffless blood pressure detection model comprises: Obtaining training data, including single-point pulse wave signals, pulse wave characteristics, human physiological and pathological characteristics, and reference blood pressure data; Establishing a cuffless blood pressure detection model, wherein the cuffless blood pressure detection model includes a deep learning model and a theoretical formula model; The pulse wave signal, pulse wave features, human physiological characteristics and reference blood pressure data are used as inputs of a cuffless blood pressure detection model to train the cuffless blood pressure detection model, wherein the pulse wave signal, human physiological characteristics and reference blood pressure data are used as input features of a deep learning model, and the pulse wave features are used as input features of a theoretical formula model.

16. The blood pressure monitoring method according to claim 16, wherein the pulse wave characteristics are obtained by extracting a single-point pulse wave signal, including waveform characteristics and pulse wave characteristics related to the cardiovascular health of the human body, wherein: Waveform characteristics include peak height, peak width, peak position, and the slope of the peak rising or falling edge; the pulse wave characteristics related to human cardiovascular health include reflected wave conduction time RWTT, radial artery augmentation index AI r and heart rate HR.

17. The blood pressure monitoring method according to claim 17, wherein the backbone network of the deep learning model adopts a time series prediction model based on Transformer or other neural network architecture, including TimesNet, ETSformer, FEDformer, Pyraformer, Autoformer, Informer or Reformer.

18. The blood pressure monitoring method according to any one of claims 16 to 18, wherein the theoretical formula model is a theoretical formula model based on the MK equation with the reflected wave transmission time RWTT as the core, and the formula is as follows: BP = a·lnRWTT+b·lnAI r +c·HR+d In the formula, BP is the human blood pressure value, a, b, c, and d are model parameters, which are calculated by the deep learning model based on the single-point pulse wave signal, human physiological characteristics, and reference blood pressure values.

Images