A blood pressure measurement method and related apparatus

Abstract

The application discloses a blood pressure measurement method and related devices. When the electronic device measures the blood pressure of the first user, if it is detected that the first user does not belong to the first group, a first blood pressure model is trained based on a physiological signal segment of the first user to obtain a second blood pressure model. The electronic device uses the second blood pressure model to obtain the blood pressure measurement result of the first user. In this way, the electronic device can obtain a more accurate blood pressure measurement result.

Description

Technical Field

[0001] The present application relates to the field of terminals, and in particular to a blood pressure measurement method and related devices. Background Art

[0002] Currently, traditional blood pressure measurement methods require medical personnel to use specialized blood pressure measurement equipment, which is inconvenient and cannot monitor the user's blood pressure in real time. In recent years, the number of wearable devices offering blood pressure measurement capabilities has increased. However, when users use wearable devices to measure their blood pressure, the blood pressure results they obtain are not very accurate. Summary of the Invention

[0003] This application provides a blood pressure measurement method and related apparatus. When measuring the blood pressure of a first user, if it is detected that the characteristics of the first user's physiological signals have low similarity with those of a first group of people, a first blood pressure model is trained based on segments of the first user's physiological signals to obtain a second blood pressure model. The electronic device then uses the second blood pressure model to measure the first user's blood pressure, thereby obtaining a more accurate blood pressure measurement result.

[0004] In a first aspect, the present application provides a blood pressure measurement method, which is applied to an electronic device, the electronic device including a blood pressure sensor; the method comprising: receiving personalized information input by a first user, the personalized information including one or more of age, height, weight, and gender; collecting a first physiological signal segment of the first user through the blood pressure sensor; determining whether the first user belongs to a first population based on the first physiological signal segment using a first classifier; when it is determined that the first user does not belong to the first population, receiving a first blood pressure value input by the first user and collecting a second physiological signal segment of the first user through the blood pressure sensor; determining a first user classification result based on the second physiological signal segment using the first classifier, and determining a first blood pressure measurement result using a first blood pressure model; determining a first loss parameter based on the first user classification result, a preset classification result, the first blood pressure value, and the first blood pressure measurement result; adjusting parameters of the first classifier based on the first loss parameter to obtain a second classifier, and adjusting parameters of the first blood pressure model to obtain a second blood pressure model; wherein the first blood pressure model is used to identify the blood pressure measurement results of the first population, and the user classification result determined by the second classifier for the second physiological signal segment is a second user classification result, the second user classification result indicating that the first user belongs to the first population; and obtaining a second blood pressure measurement result based on the second physiological signal segment using the second blood pressure model.

[0005] In this way, when the electronic device determines that the physiological signal of the first user is low in similarity to the physiological signal of the first group, that is, the first user does not belong to the first group, the first blood pressure model is calibrated using the physiological signal of the first user, so that the first blood pressure model can learn the characteristics of the physiological signal of the first user. The electronic device can obtain a more accurate blood pressure measurement result of the first user based on the calibrated first blood pressure model, that is, the second blood pressure model. Moreover, when the electronic device determines that the first user belongs to the first group, the electronic device can quickly obtain an accurate blood pressure measurement result, not only saving the time of the electronic device outputting the blood pressure measurement result, but also ensuring the accuracy of the measurement result of the non-designated group, improving the efficiency of blood pressure measurement, and reducing the cost of blood pressure measurement.

[0006] In a possible implementation, the method further includes: when it is determined that the first user belongs to the first group, obtaining a third blood pressure measurement result based on the first physiological signal segment by the first blood pressure model. In this way, generally, the first user needs to perform device calibration when measuring blood pressure for the first time or measuring blood pressure after a long time. In the process of device calibration, the user needs to maintain a standard action such as lying or sitting for a long time according to the prompt of the wearable device, so as to obtain an accurate blood pressure detection result. The calibration process is complex and inconvenient for the user. Through the blood pressure measurement method provided in the embodiment of the present application, the electronic device recognizes that the first user belongs to the first group, and directly determines the blood pressure measurement result of the user based on the first blood pressure model. Even if the first user detects blood pressure for the first time, the calibration operation does not need to be performed, and a more accurate blood pressure measurement result can be obtained.

[0007] In a possible implementation, after obtaining the second classifier and the second blood pressure model, the method further includes: collecting a third physiological signal segment of the first user by the blood pressure sensor; determining that the first user belongs to the first group based on the third physiological signal segment by the second classifier; and obtaining a fourth blood pressure measurement result based on the third physiological signal segment by the second blood pressure model. In this way, after obtaining the second blood pressure model, when the first user measures blood pressure again, the electronic device directly obtains the blood pressure measurement result of the first user based on the second blood pressure model.

[0008] In a possible implementation, the first blood pressure model comprises a first feature extraction layer and a blood pressure prediction layer; the first classifier comprises a first feature extraction layer, a second feature extraction layer and a classification layer; the first user classification result is determined by the first classifier based on the second physiological signal segment, and the first blood pressure measurement result is determined by the first blood pressure model, specifically comprising: first feature information is determined by the first feature extraction layer based on the second physiological signal segment; the first blood pressure measurement result is determined by the blood pressure prediction layer based on the first feature information; second feature information is determined by the second feature extraction layer based on the first feature information, and the second feature information is different from the first feature information; the first user classification result is determined by the classification layer based on the second feature information.

[0009] The second feature information is more detailed than the first feature information, so that the first classifier can obtain a more accurate user classification result.

[0010] In a possible implementation, the parameters of the first classifier are adjusted based on the first loss parameter to obtain a second classifier, and the parameters of the first blood pressure model are adjusted to obtain a second blood pressure model, specifically comprising: the parameters of the classification layer are adjusted in a direction of reducing the first loss parameter, and a second loss parameter of the second feature extraction layer is determined; the parameters of the second feature extraction layer are adjusted in a direction of reducing the second loss parameter, and a third loss parameter of the first feature extraction layer is determined; the parameters of the first feature extraction layer are adjusted in a direction of increasing the third loss parameter; the parameters of the blood pressure prediction layer are adjusted in a direction of reducing the first loss parameter, and a fourth loss parameter of the first feature extraction layer is determined; the parameters of the first feature extraction layer are adjusted in a direction of reducing the fourth loss parameter.

[0011] In this way, adjusting the parameters of the first feature extraction layer in a direction of not reducing the third loss parameter can make the first blood pressure model and the first classifier confuse the physiological signal features of the first user with the physiological signal features of the specified population, i.e., learn the physiological signal features of the first user. Adjusting the parameters of the first feature extraction layer in a direction of reducing the fourth loss parameter can make the first blood pressure model be able to determine a more accurate blood pressure measurement result based on the physiological signal of the first user.

[0012] In a possible implementation, before the personalized information input by the first user is received, the method further includes: displaying a first interface, the first interface including a blood pressure measurement application icon; receiving a first input for the blood pressure measurement application icon; in response to the first input, displaying a second interface, the second interface including a first control, the second interface being used for inputting the personalized information; and receiving the personalized information input by the first user, specifically including: receiving a second input for the first control, and in response to the second input, receiving the personalized information. In this way, the electronic device prompts the first user to input the personalized information, facilitating the electronic device to collect the personalized information.

[0013] In a possible implementation, the first physiological signal segment of the first user is collected by the blood pressure sensor, specifically including: in response to the second input, the first physiological signal segment of the first user is collected by the blood pressure sensor. In this way, the electronic device collects the first physiological signal segment of the first user after determining the personalized information input by the first user.

[0014] In a possible implementation, when it is determined that the first user does not belong to the first population, the method further includes: displaying a third interface, the third interface including a second control, the third interface being used for inputting a first blood pressure value; and receiving the first blood pressure value input by the first user, specifically including: receiving a third input for the second control, and in response to the third input, receiving the first blood pressure value. In this way, the electronic device determines that the first user does not belong to the first population, prompts the user to input the first blood pressure value, and facilitates the electronic device to adjust the first blood pressure model based on the first blood pressure value.

[0015] In a possible implementation, the third interface includes first prompt information, the first prompt information being used for prompting the first user to input the first blood pressure value.

[0016] In a possible implementation, the second physiological signal segment of the first user is collected by the blood pressure sensor, specifically including: in response to the third input, the second physiological signal segment of the first user is collected by the blood pressure sensor. In this way, the electronic device collects the second physiological signal segment of the first user after receiving the first blood pressure value input by the first user.

[0017] In a possible implementation, when the first physiological signal segment of the first user is collected by the blood pressure sensor, the method further includes: displaying second prompt information, the second prompt information being used for prompting the first user to maintain a preset posture, the preset posture including supine or sitting. In this way, the user maintains the preset posture, and the blood pressure measurement result recognized by the electronic device based on the obtained physiological signal is more accurate.

[0018] In a possible implementation, the first physiological signal segment includes a photoplethysmography (PPG) signal and / or an electrocardiogram (ECG) signal.

[0019] In a possible implementation, before the personalized information of the first user is received, the method further includes: receiving the first classifier and the first blood pressure model sent by the server, the first blood pressure model being trained by the physiological signal segments of the first population, the personalized information, and the blood pressure measurement result, and the first classifier being trained by the physiological signal segments of the first population. In this way, after the server obtains the first blood pressure model and the first classifier based on the training data of the first population, the server sends the first blood pressure model and the first classifier to the electronic device, so as to facilitate the electronic device to provide the blood pressure measurement function.

[0020] In a possible implementation, the first physiological signal segment has a preset segment duration.

[0021] In a possible implementation, the first physiological signal segment of the first user is collected by the blood pressure sensor, and specifically includes: collecting the first physiological signal of the first user by the blood pressure sensor within a preset collection duration; and performing a preprocessing operation on the first physiological signal to obtain the first physiological signal segment; wherein the preprocessing operation includes one or more of filtering, slicing, and alignment operations.

[0022] In a possible implementation, the first blood pressure measurement result includes a first diastolic pressure DPB and a first systolic pressure SBP, and the first blood pressure value includes a second diastolic pressure and a second systolic pressure.

[0023] In a possible implementation, the first blood pressure measurement result includes a first diastolic pressure DPB and a first systolic pressure SBP, and the first blood pressure value includes a second diastolic pressure and a second systolic pressure.

[0024] In a possible implementation, the first blood pressure measurement result includes a first diastolic pressure DPB and a first systolic pressure SBP, and the first blood pressure value includes a second diastolic pressure and a second systolic pressure. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 A hardware structure schematic diagram of an electronic device provided by an embodiment of the present application;

[0026] Figure 2 A flowchart of a blood pressure measurement method provided by an embodiment of the present application;

[0027] Figure 3 A structure schematic diagram of a first classifier provided by an embodiment of the present application;

[0028] Figure 4 A structure schematic diagram of a blood pressure training model provided by an embodiment of the present application;

[0029] Figure 5A A flowchart of a blood pressure measurement process provided by an embodiment of the present application is shown in FIG. 1.

[0030] Figure 5B A flowchart of a blood pressure measurement process provided by an embodiment of the present application is shown in FIG. 1.

[0031] Figure 6A A flowchart of a blood pressure measurement process provided by an embodiment of the present application is shown in FIG. 1.

[0032] Figure 6B A flowchart of a blood pressure measurement process provided by an embodiment of the present application is shown in FIG. 1.

[0033] Figures 7A-7F A set of interface diagrams provided by an embodiment of the present application is shown in FIG. 1. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings. In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; in the text, "and / or" only describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which means that there are three cases of A alone, A and B together, and B alone.

[0035] Hereinafter, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, the meaning of "multiple" is two or more.

[0036] The term "user interface (UI)" in the following embodiments of the present application is a medium interface for interaction and information exchange between an application program or an operating system and a user, which realizes the conversion between the internal form of information and the form acceptable by the user. The user interface is source code written in a specific computer language such as Java, extensible markup language (XML), and the like. The interface source code is parsed, rendered, and finally presented as content recognizable by the user on the electronic device. The commonly used form of the user interface is a graphic user interface (GUI), which refers to a user interface displayed in a graphical manner related to computer operation. It can include visible interface elements such as text, icons, buttons, menus, tabs, text boxes, dialog boxes, status bars, navigation bars, Widgets, and the like displayed in the display screen of the electronic device. As a medium interface for interaction and information exchange between an application program and a user, the interface needs to be generated by the electronic device for the application program in the foreground every time a vertical synchronization signal arrives.

[0037] In a possible implementation, the server can train a first blood pressure model by using physiological signals, personalized information, and blood pressure measurement results of multiple users as training data. The first blood pressure model can be used to determine the blood pressure measurement result of a user based on the physiological signals and the personalized information of the user. The physiological signals can include, but are not limited to, photoplethysmography (PPG) signals and / or electrocardiogram (ECG) signals. The personalized information can include, but is not limited to, one or more of age, weight, height, gender, and the like. The blood pressure measurement result can include systolic blood pressure (SBP) and diastolic blood pressure (DPB). In the embodiments of the present application, the user group providing the training data constitutes a specified population.

[0038] Since the training data obtained by the server is limited and only includes data of the specified population, when the electronic device measures the blood pressure of the user by using the first blood pressure model, if the similarity between the characteristics of the physiological signals of the user and the characteristics of the physiological signals of the specified population is high, the electronic device can obtain a relatively accurate blood pressure measurement result. If the similarity between the characteristics of the physiological signals of the user and the characteristics of the physiological signals of the specified population is low, the accuracy of the blood pressure measurement result obtained by the electronic device is low, and the error is large.

[0039] Therefore, the embodiment of the present application provides a blood pressure measurement method. The electronic device can receive personalized information of a first user and collect a first physiological signal segment of the first user. The electronic device can use a first classifier to determine whether the first user belongs to a specified population based on the first physiological signal segment. The first classifier is trained by physiological signals of the specified population. When the electronic device determines that the first user belongs to the specified population, the electronic device can determine a blood pressure measurement result of the first user based on the first physiological signal segment, the personalized information, and a first blood pressure model. When the electronic device determines that the first user does not belong to the specified population, the electronic device can collect a second physiological signal segment of the first user and receive a first blood pressure value input by the first user. The electronic device can construct a blood pressure training model by using the first classifier and the first blood pressure model. The electronic device can use the blood pressure training model to adjust the first blood pressure model and the first classifier based on the second physiological signal segment, a preset classification result, and the first blood pressure value. When a user classification result output by the blood pressure training model indicates that the first user belongs to the specified population, the first blood pressure model in the blood pressure training model is adjusted to obtain a second blood pressure model. The electronic device can determine the blood pressure measurement result of the first user based on the second physiological signal segment and the personalized information of the first user by using the second blood pressure model.

[0040] In this way, the electronic device can identify whether the physiological signal of the first user is similar to the physiological signal of the specified population by using the first classifier. When the electronic device determines that the similarity between the physiological signal of the first user and the physiological signal of the specified population is low, that is, the first user does not belong to the specified population, the electronic device can calibrate the first blood pressure model by using the physiological signal of the first user, so that the first blood pressure model can learn the characteristics of the physiological signal of the first user. The electronic device can obtain a more accurate blood pressure measurement result based on the calibrated first blood pressure model, that is, the second blood pressure model. Moreover, when the electronic device determines that the first user belongs to the specified population, the electronic device can quickly obtain an accurate blood pressure measurement result. By using the blood pressure measurement method provided in the embodiment of the present application, the time for the electronic device to output the blood pressure measurement result can be saved, the accuracy of the measurement result of the non-specified population can be ensured, the efficiency of blood pressure measurement can be improved, and the cost of blood pressure measurement can be reduced.

[0041] Next, a hardware structure schematic diagram of the electronic device provided in the embodiment of the present application is introduced.

[0042] The electronic device is a device having a PPG signal and / or ECG signal acquisition function. Exemplarily, the electronic device may be an electronic device having a blood pressure sensor. The blood pressure sensor is used to acquire PPG signals and / or ECG signals. For example, the electronic device having a blood pressure sensor may be a wearable device having a blood pressure sensor (e.g., a watch, a bracelet, etc.). For example, the blood pressure sensor may be a photoelectric sensor. It should be noted that the blood pressure measurement method provided in the embodiments of the present application may be implemented in other electronic devices having a blood pressure sensor, not limited to wearable devices. For example, the electronic device may be a mobile phone, a tablet, etc. having a blood pressure sensor.

[0043] In some examples, the electronic device may establish a communication connection with a device having a PPG signal and / or ECG signal acquisition function. For example, the electronic device may establish a communication connection with a device having a PPG signal and / or ECG signal acquisition function through wireless communication or wired communication. For example, near field communication (NFC) communication connection, wireless fidelity (Wi-Fi) communication connection, ultra wide band (UWB) communication connection, Bluetooth communication connection, Zigbee communication connection, and the like. The electronic device may obtain the PPG signal and / or ECG signal from the device having the PPG signal and / or ECG signal acquisition function through the communication connection.

[0044] It should be noted that the device for collecting PPG signals and / or ECG signals can be the electronic device itself, or a device that can support the electronic device to implement this function, such as a chip system or a combination device or component that can implement the functions of the electronic device, and the device can be installed in the electronic device. The embodiments of this application do not limit the specific technology and specific device form used by the electronic device.

[0045] Next, we take an electronic device including a blood pressure sensor as an example to illustrate the hardware structure of the electronic device. Figure 1 As shown, the electronic device may include but is not limited to a processor 11, a memory 12, a display screen 13, a blood pressure sensor 14, etc.

[0046] It is understood that the structures illustrated in the embodiments of the present invention do not constitute specific limitations on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown, or may combine or separate certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0047] The processor 11 can include one or more processing units, for example: the processor 11 can include an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units can be independent devices, or can be integrated in one or more processors.

[0048] The DSP is used to process digital signals, in addition to being able to process digital image signals, it can also process other digital signals. For example, when the electronic device is in frequency point selection, the digital signal processor is used to perform Fourier transform on the frequency point energy, etc.

[0049] The controller can generate operation control signals according to instruction operation codes and timing signals, and complete the control of fetching instructions and executing instructions.

[0050] The memory can also be provided in the processor 11, used to store instructions and data. In some embodiments, the memory in the processor 11 is a cache memory. The memory can save instructions or data that the processor 11 has just used or repeatedly uses. If the processor 11 needs to use the instructions or data again, it can directly call from the memory. Avoid repeated access, reduce the waiting time of the processor 11, and thus improve the efficiency of the system.

[0051] In some embodiments, the processor 11 can include one or more interfaces. The interfaces can include an inter-integrated circuit (I2C) interface, an inter-integrated circuit sound (I2S) interface, a pulse code modulation (PCM) interface, a universal asynchronous receiver / transmitter (UART) interface, a mobile industry processor interface (MIPI), a general-purpose input / output (GPIO) interface, a subscriber identity module (SIM) interface, and / or a universal serial bus (USB) interface, etc.

[0052] The MIPI interface can be used to connect the processor 11 and peripheral devices such as the display screen 13. The MIPI interface includes a display serial interface (DSI) and the like. In some embodiments, the processor 11 and the display screen 13 communicate through the DSI interface to implement the display function of the electronic device.

[0053] The GPIO interface can be configured by software. The GPIO interface can be configured as a control signal or as a data signal. In some embodiments, the GPIO interface can be used to connect the processor 11 and a blood pressure sensor 14 and the like. The GPIO interface can also be configured as an I2C interface, an I2S interface, a UART interface, a MIPI interface, and the like.

[0054] It can be understood that the interface connection relationship between the modules shown in the embodiments of the present application is only illustrative and does not constitute a structural limitation on the electronic device. In some other embodiments of the present application, the electronic device can also use different interface connection methods or a combination of multiple interface connection methods.

[0055] The memory 12 can include one or more random access memories (RAMs) and one or more non-volatile memories (NVMs). The random access memory can include static random-access memory (SRAM), dynamic random-access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM, such as the fifth generation DDR SDRAM commonly referred to as DDR5 SDRAM), and the like; the non-volatile memory can include magnetic disk storage devices, flash memories. The flash memories can include NOR FLASH, NAND FLASH, 3D NAND FLASH, and the like according to operating principles, and can include single-level cell (SLC), multi-level cell (MLC), triple-level cell (TLC), quad-level cell (QLC), and the like according to storage cell potential order, and can include universal flash storage (UFS), embedded multi media Card (eMMC), and the like according to storage specifications. The random access memory can be directly read and written by the processor 11, and can be used to store executable programs (such as machine instructions) of an operating system or other programs running, and can also be used to store data of users and application programs, and the like. The non-volatile memory can also store executable programs and store data of users and application programs, and the like, and can be loaded into the random access memory in advance for direct reading and writing by the processor 11.

[0056] The electronic device implements display functions through the GPU, the display screen 13, and the application processor, and the like. The GPU is a microprocessor for image processing, connected to the display screen 13 and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor 11 can include one or more GPUs that execute program instructions to generate or change display information.

[0057] The display screen 13 is configured to display images, videos, and the like. The display screen 13 includes a display panel. The display panel can be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode (AMOLED), a flex light-emitting diode (FLED), a Miniled, a MicroLed, a Micro-oLed, a quantum dot light emitting diodes (QLED), or the like. In some embodiments, the electronic device can include one or N display screens 13, where N is a positive integer greater than 1.

[0058] The blood pressure sensor 14 can be configured to collect physiological signals of a user, which can include one or more of PPG signals, ECG signals, and the like. For example, the blood pressure sensor 14 can be an optical sensor. In some embodiments, the electronic device can include one or N blood pressure sensors 14, where N is a positive integer greater than 1.

[0059] In the embodiments of the present application, the electronic device can implement a blood pressure measurement function through the processor 11, the blood pressure sensor 14, the display screen 13, and the like. Specifically, the electronic device needs to touch a body part (e.g., a wrist, a finger, or the like) of a user through the blood pressure sensor 14. The blood pressure sensor 14 can collect physiological signals of the user when in contact with the body part of the user, and send the collected physiological signals to the processor 11. The processor 11 can determine a blood pressure measurement result of the user based on the personalized information of the user and the physiological signals collected by the blood pressure sensor 14. The display screen 13 can be configured to display the blood pressure measurement result determined by the electronic device.

[0060] Next, a flowchart of a blood pressure measurement method provided by the embodiments of the present application is introduced.

[0061] As shown in the example of FIG. 7, the blood pressure measurement method includes the following steps: Figure 2

[0062] S201. The electronic device collects a first physiological signal segment and personalized information of a first user.

[0063] ​Specifically, the electronic device can collect a first physiological signal segment and personalized information of the first user in response to the input received for the blood pressure measurement application. The first physiological signal segment can include one or more of a PPG signal, an ECG signal, etc. The personalized information can include one or more of gender, age, height, weight, etc. The electronic device can collect the first physiological signal segment through the blood pressure sensor within a preset segment duration (e.g., 10 seconds) after receiving the input for the blood pressure measurement application.

[0064] In some examples, the first physiological signal segment collected by the electronic device includes a PPG signal collected by the electronic device at one part of the user. For example, the electronic device can collect the PPG signal at any part of the user (e.g., a finger, a wrist, etc.) through the blood pressure sensor to obtain the first physiological signal segment. In other examples, the first physiological signal segment collected by the electronic device includes PPG signals collected by the electronic device at two different parts of the user. In other examples, the first physiological signal segment collected by the electronic device includes a PPG signal and an ECG signal.

[0065] In some examples, the electronic device can detect whether the quality of the first physiological signal segment meets a signal quality requirement after obtaining the first physiological signal segment. When the electronic device detects that the quality of the first physiological signal segment meets the signal quality requirement, the electronic device can send the first physiological signal segment to the first classifier. When the electronic device detects that the quality of the first physiological signal segment does not meet the signal quality requirement, the electronic device can re-collect the first physiological signal. The electronic device can continue to collect the first physiological signal until the quality of the collected first physiological signal segment meets the signal quality requirement. For example, the electronic device can detect whether the quality of the first physiological signal segment meets the signal quality requirement based on a quality index feature of the signal (e.g., amplitude, peak, trough, signal-to-noise ratio, etc.). For another example, the electronic device can also detect whether the first physiological signal segment has a signal missing condition, etc. In this way, the electronic device can evaluate the quality of the preprocessed first physiological signal segment to ensure that the first physiological signal segment is reliable and accurate enough.

[0066] In some examples, the electronic device can output a collection prompt before collecting the first physiological signal segment. The collection prompt can be used to prompt the user to maintain a preset posture during blood pressure measurement. The preset posture can include, but is not limited to, supine, sitting, etc. In this way, the electronic device can collect more accurate physiological signals when the user maintains the preset posture.

[0067] In some examples, the electronic device stores personalized information of the first user, and the electronic device can use the stored personalized information. In this way, the electronic device does not need the first user to input the personalized information, reducing the operation of the first user.

[0068] S202. The electronic device determines, by the first classifier, whether the first user belongs to a specified population based on the first physiological signal segment.

[0069] The electronic device can use the first physiological signal segment as input data of the first classifier to determine the user classification result of whether the first user belongs to the specified population. The first physiological signal segment can be used to represent the physiological characteristics of the user. In this way, when the electronic device determines that the first user belongs to the specified population by the first classifier, it indicates that the physiological signal of the first user has a high similarity with the physiological signal of the specified population. The electronic device can execute step S203 to use the first blood pressure model trained by the physiological signal of the specified population. When the electronic device determines that the first user does not belong to the specified population by the first classifier, it indicates that the physiological signal of the first user has a low similarity with the physiological signal of the specified population. The electronic device cannot directly use the first blood pressure model trained by the physiological signal of the specified population, and needs to calibrate the first blood pressure model based on the physiological signal of the first user, so that the first blood pressure model can learn the characteristics of the physiological signal of the first user, and the calibrated first blood pressure model can obtain a more accurate blood pressure measurement result of the first user.

[0070] The first blood pressure model can include but is not limited to a first feature extraction layer and a blood pressure prediction layer. The first feature extraction layer can be used to extract feature information of the physiological signal segment and the personalized information. The blood pressure prediction layer can be used to predict the blood pressure measurement result based on the feature information. The first feature extraction layer can include but is not limited to a convolution (CONV) layer, a batch normalization (BN) layer, and a rectified linear unit (ReLu).

[0071] It can be understood that the server can train the first blood pressure model based on the physiological signal of the specified population, the personalized information of the specified population, and the blood pressure measurement result of the specified population. The server can send the first blood pressure model to the electronic device after obtaining the trained first blood pressure model.

[0072] The first classifier may include, but is not limited to, one or more feature extraction layers and a classification layer. The one or more feature extraction layers may be used to extract features of the first physiological signal segment, and the one or more feature extraction layers may include, but are not limited to, a first feature extraction layer and a second feature extraction layer. It should be noted that the first feature extraction layer of the first classifier is the same as the first feature extraction layer of the first blood pressure model. The classification layer may be used to determine a user classification result based on the feature information extracted by the one or more feature extraction layers.

[0073] like Figure 3 As shown, the second feature extraction layer may include, but is not limited to, a convolutional layer, a batch normalization layer, and a linear rectifier unit. The classification layer may include, but is not limited to, a fully connected (FC) layer and a normalized exponential function (softmax) layer. Specifically, the electronic device may input the first physiological signal segment into the first feature extraction layer. The first feature extraction layer may obtain first feature information and input the first feature information into the convolutional layer of the second feature extraction layer. The electronic device may perform a convolution operation on the first feature information through the convolutional layer to extract local features of the first feature information. The electronic device may perform batch normalization on the local features output by the convolutional layer through the batch normalization layer to obtain batch normalized data. The electronic device may process the batch normalized data output by the batch normalization layer through the linear rectifier unit to obtain second feature information. The linear rectifier unit, also known as the excitation layer, can prevent data overfitting. The electronic device may process the second feature information output by the linear rectifier unit through the fully connected layer and the normalized exponential function layer to obtain a user classification result. The second feature information is more detailed than the first feature information, which facilitates the classification layer in determining the user classification result.

[0074] In some examples, the electronic device may use a binary representation of the user classification result. For example, when the value of the user classification result obtained by the electronic device is "0," it indicates that the first user providing the first physiological signal segment belongs to a specified group of people. When the value of the user classification result obtained by the electronic device is "1," it indicates that the first user providing the first physiological signal segment does not belong to the specified group of people. It should be noted that the values ​​of the user classification results are only examples and should not constitute a limitation.

[0075] It is understood that the server can train the first classifier based on the physiological signals of a specified population. After obtaining the first classifier, the server can send the first classifier to the electronic device. Here, the electronic device can use the first physiological signal segment as input data and use the first classifier to determine whether the first user providing the first physiological signal segment belongs to the specified population.

[0076] In some examples, when the data volume of the physiological data of the specified population is large, the server can set N second feature extraction layers in the first classifier, so that the server can extract feature information in the output result of the first feature extraction layer N times through the second feature extraction layer, N being a positive integer, to obtain more in-depth feature information. In this way, the server can increase the number of layers of the feature extraction layer of the first classifier, so that the first classifier can process a large amount of input data.

[0077] In some examples, the electronic device can continuously collect the first physiological signal within a preset collection time length (for example, 60s). Then, the electronic device can perform a preprocessing operation based on the first physiological signal to obtain a first physiological signal segment. Specifically, the electronic device can filter the first physiological signal and slice the filtered first physiological signal to obtain one or more first physiological signal segments, the first physiological signal segment being a signal segment of a preset segment time length in the first physiological signal. For example, the electronic device can use a band-pass filter, a notch filter, or the like to filter the first physiological signal to remove noise in the PPG signal. The electronic device can use a notch filter, a morphological filter, or the like to filter the first physiological signal to remove noise in the ECG signal. It should be noted that if the first physiological signal includes multiple signals, the electronic device can align the different signals according to the time points at which the electronic device collects the signals, and then obtain the first physiological signal segment based on the aligned signals. In this way, the electronic device can obtain a first physiological signal segment with better signal quality.

[0078] In some examples, after collecting the first physiological signal and obtaining a plurality of first physiological signal segments based on the first physiological signal, the electronic device can input the plurality of first physiological signal segments into the first classifier to obtain a plurality of user classification results. The electronic device can determine whether the first user belongs to the specified population based on the plurality of user classification results. For example, if the result values of the plurality of user classification results obtained by the electronic device are the same, the electronic device can determine whether the first user belongs to the specified population according to any user classification result. If the result values of the plurality of user classification results obtained by the electronic device are different, the electronic device can determine the number of user classification results corresponding to different result values, and determine whether the first user belongs to the specified population based on the user classification result with the largest number. Alternatively, the electronic device can re-collect the physiological signal segment of the first user, and use the re-collected physiological signal segment to determine whether the first user belongs to the specified population.

[0079] S203. The electronic device obtains a first blood pressure measurement result based on the first physiological signal segment and the personalized information through a first blood pressure model.

[0080] Specifically, the electronic device determines that the first user belongs to the specified group, and the electronic device can take the first physiological signal segment and the personalized information as inputs of the first blood pressure model to obtain a first blood pressure measurement result. The first blood pressure measurement result can include systolic pressure and diastolic pressure.

[0081] Optionally, when the electronic device collects multiple first physiological signal segments, the electronic device can obtain multiple blood pressure measurement results based on the multiple first physiological signal segments. The electronic device can obtain the first blood pressure measurement result based on the multiple blood pressure measurement results. For example, the electronic device can take a weighted average of the values of the multiple blood pressure measurement results to obtain the value of the first blood pressure measurement result.

[0082] S204. The electronic device collects a second physiological signal segment of the first user and receives a first blood pressure value input by the first user.

[0083] When the electronic device determines that the first user does not belong to the specified group, the electronic device can collect a second physiological signal segment of the first user. Specifically, the description of the electronic device collecting the second physiological signal segment can refer to the embodiment shown in step S201, which will not be repeated here.

[0084] The electronic device can also receive a first blood pressure value input by the first user. The first blood pressure value includes diastolic pressure and systolic pressure. Optionally, the electronic device stores the first blood pressure value of the first user, or the electronic device can obtain the first blood pressure value of the first user from other electronic devices. In this way, the electronic device can obtain the first blood pressure value measured by the first user recently, and adjust the first blood pressure model with reference to the first blood pressure value.

[0085] S205. The electronic device determines a user classification result through the first classifier based on the second physiological signal segment, and determines a blood pressure measurement result through the first blood pressure model.

[0086] S206. The electronic device determines a first loss parameter based on the user classification result, a preset classification result, the first blood pressure value and the blood pressure measurement result, adjusts parameters of the first classifier based on the first loss parameter to obtain a second classifier, adjusts parameters of the first blood pressure model to obtain a second blood pressure model, and the preset classification result indicates that the first user belongs to the specified group.

[0087] The electronic device can determine a first user classification result by the first classifier and determine a first blood pressure measurement result by the first blood pressure model based on the second physiological signal segment. The electronic device can determine a first loss parameter based on the first user classification result, the preset classification result, the first blood pressure value, and the first blood pressure measurement result. The electronic device can adjust parameters of the first classifier to obtain a second classifier and adjust parameters of the first blood pressure model to obtain a second blood pressure model based on the first loss parameter. The first blood pressure model is used to identify blood pressure measurement results of a specified population, the second classifier determines a second user classification result for the user classification result determined based on the second physiological signal segment, and the second user classification result indicates that the first user belongs to the specified population.

[0088] In some examples, the electronic device can construct a blood pressure training model with the first classifier and the first blood pressure model. The electronic device can adjust the first blood pressure model and the first classifier based on the second physiological signal segment, the preset classification result, and the first blood pressure value using the blood pressure training model. When the user classification result output by the blood pressure training model indicates that the first user belongs to the specified population, the first blood pressure model in the blood pressure training model is adjusted to obtain a second blood pressure model, and the first classifier in the blood pressure training model is adjusted to obtain a second classifier.

[0089] Specifically, in the forward propagation process, the electronic device can take the second physiological signal segment as input data of a blood pressure training model including the first classifier and the first blood pressure model to obtain a user classification result and a blood pressure measurement result.

[0090] When the user classification result indicates that the first user does not belong to the specified population, the electronic device can obtain a first loss parameter based on the user classification result, the preset classification result, the blood pressure measurement result, and the first blood pressure value. The electronic device can adjust parameters of the blood pressure training model by backpropagating the first loss parameter. That is, in the backpropagation process, the electronic device can adjust parameters of the first classifier and the first blood pressure model based on the first loss parameter.

[0091] The electronic device can repeat the forward propagation process and the backpropagation process until the electronic device obtains a user classification result indicating that the first user belongs to the specified population by the blood pressure training model. At this point, the electronic device determines that the first blood pressure model is successfully calibrated, and the calibrated first blood pressure model can be referred to as a second blood pressure model. The electronic device determines that the first classifier is successfully calibrated, and the calibrated first classifier can be referred to as a second classifier.

[0092] In some examples, the electronic device can obtain a classification loss parameter based on the user classification result and the preset classification result, and obtain a prediction loss parameter based on the blood pressure measurement result and the first blood pressure value. The electronic device can obtain the first loss parameter based on the classification loss parameter and the prediction loss parameter.

[0093] In some examples, the blood pressure training model further includes a gradient reversal layer, which can be used to reduce the difference between the features of the second physiological signal segment and the physiological signal features of the specified population, so as to confuse the input data with the physiological signal of the specified population. It should be noted that the electronic device can reduce the difference between the physiological signal features of the first user and the physiological signal features of the specified population through other neural network structures, and the embodiments of the present application do not limit this.

[0094] For example, as shown in Figure 4 The blood pressure training model can include a first feature extraction layer, a blood pressure prediction layer, a gradient reversal layer, a second feature extraction layer, and a classification layer. Among them, the first feature extraction layer and the blood pressure prediction layer can constitute a first blood pressure model, and the first feature extraction layer, the second feature extraction layer, and the classification layer can constitute a first classifier. The electronic device can input the second physiological signal segment as the input data of the blood pressure training model to the first feature extraction layer. The first feature extraction layer can output first feature information. The electronic device can input the first feature information as the input data of the second feature extraction layer and the blood pressure prediction layer. The second feature extraction layer can obtain second feature information based on the first feature information. The classification layer can determine the user classification result based on the second feature information. The blood pressure prediction layer can determine the blood pressure measurement result based on the first feature information. In this way, since the features of the second feature information are more detailed than the features of the first feature information, the user classification result identified by the classification layer based on the second feature information is more accurate than the user classification result identified by the classification layer based on the first feature information.

[0095] Among them, the electronic device can calculate a first loss parameter based on the user classification result and a preset classification result, and the blood pressure measurement result and the first blood pressure value when the user classification result indicates that the first user does not belong to the specified population. Then, the electronic device can adjust the parameters of the blood pressure training model based on the first loss parameter through the back propagation technology.

[0096] Specifically, the electronic device can adjust the parameters of the classification layer in the direction of reducing the first loss parameter, and determine a second loss parameter of the second feature extraction layer. The electronic device can adjust the parameters of the second feature extraction layer in the direction of reducing the second loss parameter, and determine a third loss parameter of the first feature extraction layer. The electronic device can adjust the parameters of the first feature extraction layer in the direction of increasing the third loss parameter. In this way, adjusting the parameters of the first feature extraction layer in the direction of not reducing the third loss parameter can make the first blood pressure model and the first classifier confuse the physiological signal features of the first user with the physiological signal features of the specified population, i.e. learn the physiological signal features of the first user.

[0097] The electronic device can adjust the parameters of the blood pressure prediction layer toward a direction of a smaller first loss parameter, and determine a fourth loss parameter of the first feature extraction layer. The electronic device can adjust the parameters of the first feature extraction layer toward a direction of reducing the fourth loss parameter. In this way, adjusting the parameters of the first feature extraction layer toward a direction of reducing the fourth loss parameter can enable the first blood pressure model to determine a more accurate blood pressure measurement result based on the physiological signal of the first user.

[0098] It can be understood that the gradient inversion layer only inverts the gradient value during back propagation. The gradient inversion layer does not invert the gradient value during forward propagation. For example, the gradient inversion layer can be used to multiply the gradient by a negative factor (e.g., -1) during back propagation, so that the gradient value is inverted, thereby achieving gradient subtraction.

[0099] The electronic device can use the blood pressure training model to determine a user classification result and a blood pressure measurement result based on the second physiological signal segment after adjusting the parameters of the first classifier and the first blood pressure model. When the user classification result indicates that the first user does not belong to the specified population, the electronic device adjusts the first classifier and the first blood pressure model based on the user classification result and the blood pressure measurement result, and iterates the above process until the user classification result obtained by the electronic device through the blood pressure training model indicates that the first user belongs to the specified population, and the electronic device obtains the second classifier and the second blood pressure model.

[0100] Optionally, the electronic device can use the second physiological signal segment and the personalized information of the user as input data of the blood pressure training model. In this way, the electronic device can learn more feature information of the user, so that the electronic device can obtain a more accurate blood pressure measurement result.

[0101] In some examples, the electronic device can collect multiple (e.g., 6) second physiological signal segments of the first user, and train the second classifier and the second blood pressure model using the multiple second physiological signal segments. In this way, the electronic device can learn more feature information of the first user during the training process by training the second classifier and the second blood pressure model using multiple physiological signal segments.

[0102] S207. The electronic device obtains a second blood pressure measurement result based on the second physiological signal segment and the personalized information through the second blood pressure model.

[0103] After obtaining the second blood pressure model, the electronic device may input the second physiological signal segment and the personalized information into the second blood pressure model and output a second blood pressure measurement result, which includes systolic pressure and diastolic pressure. After obtaining the second blood pressure measurement result, the electronic device may display the second blood pressure measurement result.

[0104] In some examples, the electronic device may not collect the second physiological signal segment, but may directly use one or more first physiological signal segments to train a second classifier and a second blood pressure model. The electronic device may use the second blood pressure model to obtain a second blood pressure measurement result based on the first physiological signal segment and personalized information. This reduces the time spent collecting physiological signals.

[0105] In some examples, after obtaining the second blood pressure model, the electronic device may collect a third physiological signal segment of the first user. The electronic device may use the second blood pressure model, based on the third physiological signal segment and the personalized information, to obtain a second blood pressure measurement result. This allows the electronic device to obtain a more real-time blood pressure measurement result.

[0106] It is understandable that after the electronic device obtains the second blood pressure model, after receiving the input of the first user's blood pressure measurement, the electronic device can collect the first user's fourth physiological signal in response to the input. The electronic device can use the second classifier to determine that the first user belongs to a specified group based on the fourth physiological signal. Thereafter, the electronic device can use the second blood pressure model to obtain a third blood pressure measurement result based on the fourth physiological signal segment and the first user's personalized information. In this way, because the second classifier and the second blood pressure model in the electronic device have learned the characteristics of the first user's physiological signal, an accurate blood pressure measurement result for the first user can be quickly obtained.

[0107] Next, the blood pressure measurement method provided in the embodiment of the present application is described with reference to examples.

[0108] For example, Figure 5A As shown, the electronic device captures a first physiological signal segment 1 from subject 1. Based on the first physiological signal segment 1, the electronic device determines a user classification result 1 using a first classifier. Here, the user classification result 1 determined by the electronic device indicates that subject 1 belongs to a specified population, and the electronic device does not calibrate the first blood pressure model. The electronic device can obtain a blood pressure measurement result for subject 1 based on the first physiological signal segment 1, subject 1's personalized information, and the first blood pressure model.

[0109] The electronic device collects a first physiological signal segment 2 of the subject 2. The electronic device determines a user classification result 2 by using the first classifier based on the first physiological signal segment 2. In this case, the user classification result 2 determined by the electronic device indicates that the subject 2 does not belong to the specified population, and the electronic device calibrates the first blood pressure model. The electronic device can collect a second physiological signal segment of the subject 2. The electronic device can obtain a second blood pressure model based on the second physiological signal segment, the first classifier, and the first blood pressure model. For details, please refer to Figure 2 The electronic device can use the second blood pressure model to obtain a blood pressure measurement result of the subject 2 based on the second physiological signal segment and the personalized information.

[0110] For example, the process of measuring blood pressure of a user by using the blood pressure model is as shown in Figure 5B The electronic device collects a physiological signal of a user, and pre-processes the physiological signal to obtain a physiological signal segment. The electronic device can also receive personalized information input by the user. The electronic device can input the physiological signal segment and the personalized information into the blood pressure model to obtain a blood pressure measurement result, which includes SBP and DBP. When the electronic device determines that the user belongs to the specified population based on the physiological signal, the blood pressure model is the first blood pressure model. When the electronic device determines that the user does not belong to the specified population based on the physiological signal, the blood pressure model is the second blood pressure model.

[0111] In this way, since the electronic device can determine a feature of the physiological signal of the user to obtain a determination result, and determine whether the blood pressure measurement needs to be calibrated according to the determination result, the time for outputting the result can be saved, and the accuracy of the measurement result of the non-specified population can be ensured, so as to improve the processing efficiency of the blood pressure measurement and reduce the cost of the blood pressure measurement.

[0112] Next, a training process of a first blood pressure model provided in an embodiment of the present application is introduced.

[0113] For example, as shown in Figure 6A The server can obtain training data, which includes a physiological signal of a user, personalized information, and a blood pressure measurement result. The server can train a first blood pressure model by using a deep learning algorithm based on the training data. The server can send the first blood pressure model to the electronic device after obtaining the first blood pressure model.

[0114] For example, as shown in Figure 6BAs shown, the database of the server stores training data, which includes the correspondence between physiological signals, personalized information, and blood pressure detection results. The electronic device can obtain the physiological signals and personalized information from the database. The server can preprocess the physiological signals to obtain physiological signal segments. The server can use the physiological signal segments and the personalized information as input data of the first blood pressure model to obtain output data, which includes systolic pressure and diastolic pressure. The server can obtain the blood pressure measurement results corresponding to the physiological signals and personalized information from the database, and determine the first loss parameter based on the blood pressure measurement results and the output data. The server can adjust the parameters of the first blood pressure model based on the first loss parameter until the server obtains the first blood pressure model with a preset accuracy (e.g., 90%). For a description of the first blood pressure model, please refer to Figure 2 The embodiments shown will not be described here.

[0115] It can be understood that the user who provides the training data in the embodiments of the present application belongs to a specified population. In some examples, the server can receive input training data, or the server can use crowdsourcing technology to obtain training data. The crowdsourcing technology is a method of obtaining resources through a large number of electronic devices with the permission of the user. The server can use crowdsourcing technology to collect training data through the user's electronic device without the user's awareness after obtaining the user's permission.

[0116] In some examples, the server can send the updated first blood pressure model to the electronic device every preset time period (e.g., one month). In this way, the server can continuously obtain training data and update the first blood pressure model, so that the electronic device can more accurately obtain the blood pressure measurement result.

[0117] Next, the blood pressure measurement method provided by the embodiments of the present application will be introduced in combination with application scenarios.

[0118] In a possible implementation, after receiving an input for a blood pressure measurement application icon, the electronic device can collect a first physiological signal segment of a first user and obtain personalized information of the first user in response to the input. When the first classifier is used to determine that the first user belongs to a specified population based on the first physiological signal segment, the electronic device can use the first blood pressure model to determine the blood pressure measurement result of the first user based on the physiological signals and the personalized data of the first user. When the first classifier is used to determine that the first user does not belong to the specified population based on the first physiological signal segment, the electronic device can prompt the first user to input a first blood pressure value and collect a second physiological signal segment. The electronic device can calibrate the first blood pressure model based on the first blood pressure value, a preset classification result, and the second physiological signal segment to obtain a second blood pressure model. For details, please refer to Figure 2The illustrated embodiment will not be further described here. The electronic device can use the second blood pressure model to determine the first user's blood pressure measurement result based on the second physiological signal segment and the personalized data. The electronic device can also display the first user's blood pressure measurement result. In this way, the electronic device can display a corresponding user interface during blood pressure measurement, prompting the user to perform corresponding operations, thereby facilitating the electronic device to quickly obtain accurate blood pressure measurement results.

[0119] In some examples, the electronic device may display a collection prompt when collecting a user's physiological signals. The collection prompt may be used to prompt the user to remain still or maintain a preset posture. In this way, the electronic device can more efficiently collect physiological signals with the user's cooperation. Optionally, the electronic device may display the collection prompt only when collecting physiological signals for calibrating the first blood pressure model. In this way, the electronic device can prompt the user to remain still or maintain a preset posture when calibrating the first blood pressure model, allowing the electronic device to calibrate the first blood pressure model more quickly.

[0120] Next, taking an electronic device such as a watch or a bracelet as an example, the blood pressure measurement method provided in an embodiment of the present application is introduced.

[0121] For example, Figure 7A As shown. The electronic device may display an interface 800. The interface 800 may include one or more application icons. The one or more application icons may include a blood pressure measurement application icon 801. The blood pressure measurement application icon 801 may be used to trigger the electronic device to display an interface of a blood pressure measurement application.

[0122] The electronic device may receive the first user's Figure 7A After input (e.g., single click) of the blood pressure measurement application icon 801 shown in FIG. 8 , the following is displayed in response to the input. Figure 7B The interface 810 shown. Figure 7B As shown, the interface 810 includes a user information bar 812 and a confirmation control 813. The user information bar 812 is used to input the user's personalized information, for example, one or more of gender, height, weight, age and other information.

[0123] For example, the user information bar 812 includes weight information and age information. The electronic device can receive input from the first user on the touch screen of the electronic device to type in personalized information. Alternatively, the electronic device can receive personalized information input by the first user through voice. Alternatively, the electronic device stores the personalized information of the first user, and the electronic device can display the stored personalized information in the user information bar 812. In some examples, the electronic device can obtain the weight information of the first user through a weight measuring device. Here, the user information bar 812 displays the weight information typed in by the first user: "52 kg" and age information: "23 years old". Among them, the confirmation control 813 can be used to trigger the electronic device to determine the personalized information.

[0124] Optionally, the interface 810 may further include a prompt message 811, which may be used to prompt the user to enter personalized information. The prompt message 811 may be one or more of text information, image information, and voice information. For example, the prompt message 811 may be a text prompt message: "Please enter personal information."

[0125] The electronic device can receive Figure 7B After the input of the confirmation control 813 is entered, the following is displayed in response to the input: Figure 7C The interface 820 shown. Figure 7C As shown, interface 820 may include, but is not limited to, prompt information 821. Prompt information 821 may be used to prompt the user that the electronic device is measuring blood pressure. Prompt information 821 may include, but is not limited to, text prompt information, image prompt information, animation prompt information, and voice prompt information, etc. For example, prompt information 821 may be a text prompt information: "Measuring..." Optionally, prompt information 821 may also be used to prompt the user to maintain a preset posture. In this way, when the user maintains the preset posture, it is easier for the electronic device to collect more accurate physiological signals.

[0126] The electronic device can receive Figure 7B After inputting the confirmation control 813 shown, the first physiological signal segment of the first user is collected. When the electronic device uses the first classifier to determine that the first user belongs to the specified group based on the first physiological signal segment, it can use the first blood pressure model to determine the blood pressure measurement result based on the first physiological signal segment of the first user. For details, the description of how the electronic device determines the blood pressure measurement result can be found in Figure 2 The embodiment shown is not described in detail here. Here, the diastolic pressure in the blood pressure measurement result obtained by the electronic device is 110mHg and the systolic pressure is 75mHg.

[0127] After determining the user's blood pressure measurement result, the electronic device can display the following information: Figure 7D The interface 830 shown.Figure 7D As shown, interface 830 includes result information 831, which is used to indicate the user's blood pressure measurement result. Result information 831 can be one or more of text information, image information, and voice information. For example, result information 831 can include text information such as "diastolic pressure: 110 mHg" and text information such as "systolic pressure: 75 mHg."

[0128] The electronic device can receive Figure 7B After the input of the confirmation control 813 shown in FIG. 8 is entered, the first physiological signal segment of the first user is collected. When the electronic device uses the first classifier to determine that the first user does not belong to the specified group based on the first physiological signal segment, the display is as follows: Figure 7E Interface 850 is shown.

[0129] like Figure 7E As shown, interface 850 can be used for the first user to input the first blood pressure value. Interface 850 may include but is not limited to a diastolic pressure input box 851, a systolic pressure input box 852 and a confirmation control 853. Among them, the diastolic pressure input box 851 can be used to input the diastolic pressure in the first blood pressure value. The systolic pressure input box 852 can be used to input the systolic pressure in the first blood pressure value. The confirmation control 853 can be used to trigger the electronic device to determine the first blood pressure value. Optionally, interface 850 may also include a prompt message for prompting the user to input the first blood pressure value. For example, the prompt message may be a text prompt message: "Please enter the result of your most recent blood pressure measurement."

[0130] After receiving the first user's input to the confirmation control 853, the electronic device can collect the first user's second physiological signal segment in response to the input. After collecting the second physiological signal segment, the electronic device can calibrate the first blood pressure model based on the second physiological signal segment, the first blood pressure value and the preset classification result to obtain a second blood pressure model. The electronic device can use the second blood pressure model to determine the first user's blood pressure measurement result based on the second physiological signal segment. For details, please refer to Figure 2 The electronic device can display the blood pressure measurement result as follows: Figure 7D Interface 830 is shown.

[0131] Optionally, the electronic device may display the following after receiving the input for the confirmation control 853: Figure 7F The interface 860 shown, or, as shown Figure 7C The interface 820 shown. For example, Figure 7FAs shown, the interface 860 may include but is not limited to prompt information 861. Prompt information 861 can be used to prompt the user that the electronic device is detecting blood pressure. Prompt information 861 may include but is not limited to text prompt information, image prompt information, animation prompt information, and voice prompt information, etc. For example, the prompt information 861 can be a text prompt information: "It is detected that your blood pressure is being measured for the first time. You need to measure again. Please wait patiently..." The electronic device displays as shown. Figure 7F As shown in the interface 860, the electronic device can display the following when the blood pressure measurement result is determined: Figure 7D In this way, the electronic device needs to spend more time to re-collect the physiological signals of the user, and the electronic device displays the following Figure 7F The interface 860 shown reminds the user that it will take longer time to detect blood pressure.

[0132] In some examples, the electronic device stores personalized information of the first user. Figure 7A After inputting the blood pressure measurement application icon 801, the display Figure 7C The interface 820 shown in FIG. 1 is displayed, and the first physiological signal segment of the first user is collected. Afterwards, the electronic device can determine the blood pressure measurement result of the user using the blood pressure measurement method provided in the embodiment of the present application. For details, please refer to the above embodiment and will not be repeated here.

[0133] In this way, the electronic device can implement the blood pressure measurement method provided in the embodiment of the present application through the above interface.

[0134] As described above, 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 above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of blood pressure measurement, characterized by, The method is applied to an electronic device, and the electronic device comprises a blood pressure sensor. Receiving personalized information input by a first user, the personalized information comprising one or more of age, height, weight, and gender; Collecting a first physiological signal segment of the first user by the blood pressure sensor; Determining, by a first classifier, whether the first user belongs to a first population based on the first physiological signal segment; When it is determined that the first user does not belong to the first population, receiving a first blood pressure value input by the first user, and collecting a second physiological signal segment of the first user by the blood pressure sensor; Determining, by the first classifier, a first user classification result based on the second physiological signal segment, and determining a first blood pressure measurement result by a first blood pressure model; Determining a first loss parameter based on the first user classification result, a preset classification result, the first blood pressure value, and the first blood pressure measurement result; Adjusting parameters of the first classifier based on the first loss parameter to obtain a second classifier, and adjusting parameters of the first blood pressure model to obtain a second blood pressure model; wherein the first blood pressure model is used to identify blood pressure measurement results of the first population, the second classifier determines a second user classification result for the user classification result determined based on the second physiological signal segment, and the second user classification result indicates that the first user belongs to the first population; Obtaining a second blood pressure measurement result by the second blood pressure model based on the second physiological signal segment and the personalized information.

2. The method of claim 1, wherein, The method further comprises: When it is determined that the first user belongs to the first population, obtaining a third blood pressure measurement result by the first blood pressure model based on the first physiological signal segment.

3. The method of claim 1, wherein, After obtaining the second classifier and the second blood pressure model, the method further comprises: Collecting a third physiological signal segment of the first user by the blood pressure sensor; Determining, by the second classifier, that the first user belongs to the first population based on the third physiological signal segment; Obtaining a fourth blood pressure measurement result by the second blood pressure model based on the third physiological signal segment.

4. The method of claim 1, wherein, The first blood pressure model comprises a first feature extraction layer and a blood pressure prediction layer; the first classifier comprises the first feature extraction layer, a second feature extraction layer, and a classification layer; and the determining, by the first classifier, a first user classification result based on the second physiological signal segment, and the determining, by the first blood pressure model, a first blood pressure measurement result, specifically comprises: Determining first feature information based on the second physiological signal segment by the first feature extraction layer; Determining the first blood pressure measurement result based on the first feature information by the blood pressure prediction layer; Determining second feature information based on the first feature information by the second feature extraction layer, the second feature information being different from the first feature information; Determining the first user classification result based on the second feature information by the classification layer.

5. The method of claim 4, wherein, The method further comprises: The classification layer adjusts the parameters of the classification layer in a direction of reducing the first loss parameter, and determines a second loss parameter of the second feature extraction layer; The second feature extraction layer adjusts the parameters of the second feature extraction layer in a direction of reducing the second loss parameter, and determines a third loss parameter of the first feature extraction layer; The first feature extraction layer adjusts the parameters of the first feature extraction layer in a direction of increasing the third loss parameter; The blood pressure prediction layer adjusts the parameters of the blood pressure prediction layer in a direction of reducing the first loss parameter, and determines a fourth loss parameter of the first feature extraction layer; The first feature extraction layer adjusts the parameters of the first feature extraction layer in a direction of reducing the fourth loss parameter.

6. The method of claim 1 or 2, wherein, Before the receiving of the first user input of the personalized information, the method further comprises: displaying a first interface, the first interface comprising a blood pressure measurement application icon; receiving a first input for the blood pressure measurement application icon; in response to the first input, displaying a second interface, the second interface comprising a first control, the second interface being used for inputting the personalized information; The receiving of the first user input of the personalized information specifically comprises: receiving a second input for the first control, and in response to the second input, receiving the personalized information.

7. The method of claim 6, wherein, The collecting of the first physiological signal segment of the first user by the blood pressure sensor specifically comprises: in response to the second input, collecting the first physiological signal segment by the blood pressure sensor.

8. The method of claim 7, wherein, When it is determined that the first user does not belong to the first population, the method further comprises: displaying a third interface, the third interface comprising a second control, the third interface being used for inputting the first blood pressure value; The receiving of the first user input of the first blood pressure value specifically comprises: receiving a third input for the second control, and in response to the third input, receiving the first blood pressure value.

9. The method of claim 8, wherein, The third interface comprises first prompt information, the first prompt information being used for prompting the first user to input the first blood pressure value.

10. The method of claim 8, wherein, The collecting of the second physiological signal segment of the first user by the blood pressure sensor specifically comprises: in response to the third input, collecting the second physiological signal segment by the blood pressure sensor.

11. The method of claim 7, wherein, When the first physiological signal segment of the first user is collected by the blood pressure sensor, the method further comprises: displaying second prompt information, the second prompt information being used for prompting the first user to maintain a preset posture, the preset posture comprising supine or sitting.

12. The method of claim 1, wherein, The first physiological signal segment comprises a photoplethysmogram (PPG) signal and / or an electrocardiogram (ECG) signal.

13. The method of claim 1, wherein, Before the receiving of the first user input of the personalized information, the method further comprises: receive the first classifier and the first blood pressure model sent by the server, the first blood pressure model being trained by physiological signal segments, personalized information and blood pressure measurement results of the first population, and the first classifier being trained by physiological signal segments of the first population.

14. The method of claim 1, wherein, The first physiological signal segment has a preset segment duration.

15. The method of claim 14, wherein, The first physiological signal segment of the first user is collected by the blood pressure sensor, specifically including: The first physiological signal of the first user is collected by the blood pressure sensor within a preset collection duration. The first physiological signal is preprocessed to obtain the first physiological signal segment, wherein the preprocessing operation includes one or more of filtering, slicing and alignment operation.

16. The method of claim 1, wherein, The first blood pressure measurement result includes a first diastolic pressure DPB and a first systolic pressure SBP, and the first blood pressure value includes a second diastolic pressure and a second systolic pressure.

17. An electronic device, characterized in that: including: a blood pressure sensor, one or more processors and one or more memories; The blood pressure sensor, the one or more memories and the one or more processors are coupled, and the one or more memories are used to store executable programs, so that the electronic device executes the method as claimed in any one of claims 1-16 when the one or more processors execute the executable programs.

18. A readable storage medium storing a program, characterized in that, When the program runs on the electronic device, the electronic device executes the method as claimed in any one of claims 1-16.

Images