Intelligent blood pressure measuring method and system based on deflation type oscillography

Abstract

The invention relates to an intelligent blood pressure measuring method and system based on deflation oscillography, and belongs to the technical field of medical detection.The measuring method comprises the steps that an air pump is controlled to inflate air into a cuff of the wrist or the upper arm of a testee till a preset initial air pressure value is reached; controlling an electromagnetic valve to release gas in the cuff at a preset rate, and receiving air pressure data and oscillatory wave signals, which are collected by a pressure sensor in real time, in the cuff; identifying the oscillatory wave signal based on a characteristic value method and extracting key characteristics to obtain oscillatory wave characteristic parameters; based on a pre-established proportionality coefficient model, a preliminary blood pressure value is calculated according to the air pressure data and the oscillatory wave characteristic parameters; and inputting the oscillatory wave signal, the oscillatory wave characteristic parameter and the initial blood pressure value into a pre-trained blood pressure value prediction model with a momentum item to obtain a final blood pressure value, and sending the final blood pressure value to a display module for output and display. The precision and stability of blood pressure measurement are improved.

Description

Technical Field

[0001] The present application relates to the field of medical detection technology, and in particular to an intelligent blood pressure measurement method and system based on a deflation oscillometric method. Background Art

[0002] Blood pressure is one of the important physiological parameters for assessing cardiovascular health. Accurate and reliable blood pressure measurement is crucial for early diagnosis of diseases, monitoring of treatment effects, and personal health management. Traditional methods of blood pressure measurement mainly include auscultation and oscillometric methods. Among them, the auscultation method determines systolic and diastolic blood pressure by doctors using a stethoscope to listen to Korotkoff sounds. Although this method is widely used in clinical practice, it has some significant limitations. In contrast, the oscillometric method, as a non-invasive automated blood pressure measurement technology, has been widely used because of its simplicity and ease of use. The oscillometric method uses the changes in air pressure in the cuff to detect the oscillation wave signals caused by the pulse, and calculates the blood pressure value by analyzing these signals.

[0003] However, due to the influence of environmental factors and individual differences, the oscillometric measurement method still has problems such as insufficient accuracy and susceptibility to interference in signal processing and blood pressure value calculation. Especially when facing complex or abnormal physiological conditions, such as arrhythmia, changes in vascular elasticity, etc., it is easy for the oscillometric sphygmomanometer to have large fluctuations in the results of multiple measurements, making it difficult to provide stable and reliable blood pressure readings. Summary of the invention

[0004] In order to improve the accuracy and stability of blood pressure measurement, the present application provides an intelligent blood pressure measurement method and system based on the deflation oscillometric method.

[0005] In a first aspect, the present application provides an intelligent blood pressure measurement method based on the deflation oscillometric method, which adopts the following technical solutions: An intelligent blood pressure measurement method based on deflation oscillometric method is applied to an intelligent blood pressure measurement system including an air pump, a solenoid valve, a pressure sensor, a microprocessor, a display module and a cuff, and the measurement method includes: Controlling the air pump to inflate air into the cuff on the wrist or upper arm of the subject until a preset initial air pressure value is reached; Controlling the solenoid valve to release the gas in the cuff at a preset rate, and receiving the air pressure data and oscillation wave signal in the cuff collected in real time by the pressure sensor; Identify the oscillation wave signal and extract key features based on the eigenvalue method to obtain oscillation wave characteristic parameters; Based on a pre-established proportional coefficient model, a preliminary blood pressure value is calculated according to the air pressure data and the oscillation wave characteristic parameters; the oscillation wave signal, the oscillation wave characteristic parameters and the preliminary blood pressure value are input into a pre-trained blood pressure value prediction model with a momentum term to obtain a final blood pressure value; wherein the blood pressure value includes a systolic pressure value and a diastolic pressure value; The final blood pressure value is sent to the display module for output display.

[0006] By adopting the above technical solutions, from air pressure control to data analysis to output display, by combining the feature processing of oscillation waves and deep learning of neural networks, the intelligent blood pressure measurement process is made more efficient and accurate. When faced with complex physiological signals, the reliability and accuracy of blood pressure measurement results can still be ensured, thereby providing users with an efficient, convenient and reliable blood pressure monitoring solution, improving user experience, and providing more accurate and stable data support for the field of medical testing.

[0007] Optionally, the step of identifying the oscillation wave signal and extracting key features based on the eigenvalue method to obtain oscillation wave characteristic parameters includes: Performing data preprocessing on the oscillation wave signal; dividing the preprocessed oscillation wave signal into a plurality of sub-segments according to different preset pressure ranges; Performing peak detection and trough detection on the oscillation wave signal in each sub-segment, determining the peak position and the trough position, and calculating the characteristic parameters between each peak and trough to obtain the key characteristic parameters of the oscillation wave signal in each sub-segment; Based on the principal component analysis method, dimension reduction processing and feature selection are performed on the key characteristic parameters of the oscillation wave signal in each sub-segment, and the selected features are normalized to obtain the optimized characteristic parameters of the oscillation wave.

[0008] By adopting the above technical solutions, the system can efficiently and accurately extract the key characteristic parameters of the oscillation wave signal. These steps not only improve the accuracy and stability of feature extraction, but also reduce data redundancy and enhance the robustness and scalability of the system.

[0009] Optionally, based on a pre-established proportionality coefficient model, the step of calculating a preliminary blood pressure value according to the air pressure data and the oscillation wave characteristic parameters includes: Load a pre-built scale factor model from system storage; Formatting the air pressure data and oscillation wave characteristic parameters; Mapping the oscillation wave characteristic parameters in each sub-segment to corresponding input variables in the proportional coefficient model to obtain a mapped characteristic parameter set; Correcting the air pressure data, and time-aligning the corrected air pressure data with the characteristic parameter set to obtain an air pressure data set; The characteristic parameter set and the air pressure data set are input into a proportional coefficient model to calculate preliminary systolic pressure value and diastolic pressure value.

[0010] By adopting the above technical solution, based on the pre-established proportional coefficient relationship, the preliminary blood pressure value is calculated using the air pressure data and the oscillation wave characteristic parameters. This technical solution not only improves the accuracy and stability of the preliminary blood pressure estimation, but also provides a reliable data basis for subsequent neural network optimization, significantly improving the accuracy and reliability of the overall blood pressure measurement.

[0011] Optionally, the relationship of the proportional coefficient model is: Systolic blood pressure = a1×amplitude+b1×frequency+c1×pressure+d1; Diastolic pressure = a2×amplitude+b2×frequency+c2×pressure+d2; In the above formula, a1, b1, c1, d1 and a2, b2, c2, d2 are predetermined proportional coefficients.

[0012] Optionally, the method further includes a step of training the blood pressure value prediction model, wherein the training step includes: Acquire a sample data set and perform data preprocessing; wherein the sample data set includes oscillation wave signal sample data, oscillation wave characteristic parameter sample data, preliminary blood pressure sample values ​​and corresponding blood pressure value label data, and the blood pressure value label data is the actually measured systolic and diastolic pressure; Dividing the sample data set into a training set, a validation set and a test set; Inputting the training set into a pre-built convolutional neural network model for iterative training, calculating the loss function of the model, calculating the gradient through a back-propagation algorithm, and optimizing the model parameters; The weights of the convolutional neural network model are updated based on an optimization algorithm with a momentum term until the loss function meets a preset condition, thereby obtaining a trained convolutional neural network model; Validating the convolutional neural network model based on the validation set, evaluating the performance of the convolutional neural network model and adjusting the hyperparameters of the model according to the validation results; The prediction ability of the adjusted convolutional neural network model is tested based on the test set to obtain the blood pressure value prediction model.

[0013] By adopting the above technical solution, the convolutional neural network model is iteratively trained using an optimization algorithm with a momentum term, and the model performance is evaluated and optimized through validation sets and test sets. This technical solution not only improves the accuracy of blood pressure measurement, but also enhances the robustness of the model to abnormal situations. This optimization method significantly improves the accuracy and stability of blood pressure measurement, providing a more reliable and efficient solution for the field of medical testing.

[0014] Optionally, after the step of sending the final blood pressure value to the display module for output display, the method further includes: Determining whether the final blood pressure value exceeds a normal blood pressure range; If so, a blood pressure warning prompt is generated according to the final blood pressure value and sent to the corresponding medical terminal.

[0015] By adopting the above technical solution, after the final blood pressure value is sent to the display module for output display, it is further determined whether the blood pressure value exceeds the normal range. If it exceeds the normal range, a corresponding early warning prompt is generated and sent to the corresponding medical terminal. This technical solution can not only monitor the user's blood pressure status in real time, but also notify medical staff in time when abnormalities are found, thereby improving the response speed and quality of medical services. Through this early warning mechanism, patients can receive more timely medical intervention, reduce the risks caused by high blood pressure or low blood pressure, and improve the safety and effectiveness of overall medical care.

[0016] In a second aspect, the present application provides an intelligent blood pressure measurement system based on the deflation oscillometric method, which adopts the following technical solutions: An intelligent blood pressure measurement system based on deflation oscillometric method, the intelligent blood pressure measurement system comprising an air pump, a solenoid valve, a pressure sensor, a microprocessor, a display module and a cuff wrapped around the wrist or upper arm of a subject; The air pump is used to fill the cuff with gas; The solenoid valve is connected to the air pump and is used to control the inflow and outflow of gas in the cuff; The cuff is used to adjust the internal air pressure through the air pump and the solenoid valve; The pressure sensor is connected to the cuff and is used to collect air pressure data and oscillation wave signals in the cuff in real time; The microprocessor is connected to the air pump, the solenoid valve and the pressure sensor, and is used to receive and process the air pressure data and the oscillation wave signal collected by the pressure sensor, and calculate the final blood pressure value; The display module is used to display the final blood pressure value calculated by the microprocessor.

[0017] Optionally, the microprocessor includes: An air pump control unit, used for controlling the air pump to inflate air into the cuff of the wrist or upper arm of the subject until a preset initial air pressure value is reached; A solenoid valve control unit, used for controlling the solenoid valve to release the gas in the cuff at a preset rate; A data receiving unit, used for receiving the air pressure data and oscillation wave signal in the cuff collected in real time by the pressure sensor; A feature extraction unit, used to identify the oscillation wave signal and extract key features based on a eigenvalue method to obtain oscillation wave feature parameters; a preliminary blood pressure value generating unit, configured to calculate the preliminary blood pressure value according to the air pressure data and the oscillation wave characteristic parameters based on a pre-established proportional coefficient model; A final blood pressure value generating unit, used for inputting the oscillation wave signal, the oscillation wave characteristic parameter and the preliminary blood pressure value into a pre-trained blood pressure value prediction model with a momentum term to obtain a final blood pressure value; wherein the blood pressure value includes a systolic pressure value and a diastolic pressure value; The final blood pressure value sending unit is used to send the final blood pressure value to the display module for output display.

[0018] In a third aspect, the present application provides a computer device, which adopts the following technical solution: A computer device comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method as described in the first aspect.

[0019] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any one of the methods in the first aspect.

[0020] To summarize, the present application includes at least one of the following beneficial technical effects: from air pressure regulation to data analysis to output display, by combining the feature processing of oscillation waves and deep learning of neural networks, the efficiency and precision of the intelligent blood pressure measurement process are achieved, and the reliability and accuracy of blood pressure measurement results can still be ensured when facing complex physiological signals, thereby providing users with an efficient, convenient and reliable blood pressure monitoring solution, improving user experience, and providing more accurate and stable data support for the field of medical testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a first flow chart of an intelligent blood pressure measurement method according to one of the embodiments of the present application.

[0022] Figure 2This is a second flow chart of the intelligent blood pressure measurement method according to one of the embodiments of the present application.

[0023] Figure 3 This is a third flow chart of the intelligent blood pressure measurement method according to one of the embodiments of the present application.

[0024] Figure 4 This is a fourth flow chart of the intelligent blood pressure measurement method according to one embodiment of the present application.

[0025] Figure 5 This is a fifth flow chart of the intelligent blood pressure measurement method according to one of the embodiments of the present application. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-5 It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0027] The embodiment of the present application discloses an intelligent blood pressure measurement method based on the deflation oscillometric method.

[0028] Reference Figure 1 , an intelligent blood pressure measurement method based on deflation oscillometric method, applied to an intelligent blood pressure measurement system including an air pump, a solenoid valve, a pressure sensor, a microprocessor, a display module and a cuff, the measurement method comprising: Step S101, control the air pump to inflate the cuff of the wrist or upper arm of the subject until a preset initial air pressure value is reached; specifically, during the inflation stage, the subject sits quietly, with the arm aligned with the heart, and wears the cuff, and the microprocessor sends a command to control the air pump to inflate the cuff of the wrist or upper arm of the subject, and the air pressure inside the cuff gradually increases until a preset initial air pressure value (such as about 180 mmHg) is reached. At this time, the arterial blood flow is blocked, and the pressure change in the cuff is mainly composed of the oscillation signal caused by the pulse. By quickly inflating to a predetermined pressure level, it is ensured that the blood flow can be completely blocked and a detectable oscillation wave signal is generated, providing a basis for subsequent blood pressure measurement.

[0029] It should be noted that in order to ensure the smoothness and stability of the inflation process, the present application also introduces a momentum term parameter, which is similar to the integral term (I) and differential term (D) in the PID controller, and is used to smooth pressure changes and reduce fluctuations. For example, during the inflation stage, the target air pressure is set to 180 mmHg. Initially, the air pump works at full speed and inflates quickly. If the current air pressure is 170 mmHg and the deviation is 10 mmHg, the proportional term will increase the working intensity of the air pump; the integral term will accumulate past deviations and gradually increase the working intensity of the air pump; the differential term will be adjusted according to the speed of the air pressure rise to avoid overshoot, so as to achieve a smoother inflation process and reduce the pressure fluctuations caused by rapid inflation, thereby improving the accuracy of subsequent measurements.

[0030] Step S102, controlling the solenoid valve to release the gas in the cuff at a preset rate, and receiving the air pressure data and oscillation wave signal in the cuff collected in real time by the pressure sensor; Specifically, during the deflation stage, the microprocessor controls the solenoid valve to release the gas in the cuff at a set speed (such as 5-10 mmHg / s). At the same time, the pressure sensor continuously monitors and records the air pressure changes and oscillation wave signals in the cuff, thereby obtaining a series of air pressure data and oscillation waveform information that change over time.

[0031] It should be noted that in order to ensure the smoothness and stability of the deflation process, the present application also introduces a momentum term parameter, which is similar to the proportional term (P), integral term (I) and differential term (D) in the PID controller, and is used to adjust the deflation rate to make it more stable; exemplarily, in the deflation stage, the deflation rate is set to 5-10 mmHg / s. Initially, the solenoid valve opens at the set speed and starts to deflate. The PID controller adjusts according to the deviation between the current air pressure and the target air pressure to ensure a stable deflation rate; for example, if the current air pressure is 170 mmHg, the target air pressure is 160 mmHg, and the deviation is 10 mmHg, the proportional term will increase the degree of opening of the solenoid valve; the integral term will accumulate past deviations and gradually increase the degree of opening of the solenoid valve; the differential term will be adjusted according to the speed of the air pressure drop to avoid the air pressure dropping too fast, so as to achieve a smoother deflation process and reduce the air pressure fluctuation caused by uneven deflation rate, thereby improving the accuracy and stability of blood pressure measurement.

[0032] It can be understood that by slowly and evenly reducing the pressure in the cuff, the system is able to capture the oscillation wave characteristics under different pressures, which is the key basis for calculating blood pressure values.

[0033] Step S103, identifying the oscillation wave signal and extracting key features based on the eigenvalue method to obtain oscillation wave characteristic parameters; In some embodiments, the microprocessor uses the eigenvalue method to analyze the oscillation wave signal, identify key points therein, such as peaks and troughs, and obtain important parameters describing the oscillation wave morphology, including but not limited to amplitude, frequency, etc. By processing the original signal, the information that best reflects the blood pressure state is extracted, providing reliable data support for subsequent accurate calculation of blood pressure.

[0034] Step S104, based on a pre-established proportional coefficient model, a preliminary blood pressure value is calculated according to the air pressure data and the oscillation wave characteristic parameters; Among them, the microprocessor can use the proportional coefficient method to convert the characteristic parameters into preliminary systolic and diastolic pressure estimates based on the proportional coefficient model established by clinical data and the oscillation wave characteristic parameters obtained in the above steps, laying the foundation for further optimization.

[0035] Step S105, input the oscillation wave signal, oscillation wave characteristic parameters and preliminary blood pressure value into a pre-trained blood pressure value prediction model with momentum term to obtain the final blood pressure value; wherein the blood pressure value includes systolic pressure value and diastolic pressure value; specifically, after learning from a large number of actual blood pressure measurement data sets, the model has the ability to accurately predict blood pressure from complex signals and can achieve further optimization of the preliminary blood pressure value.

[0036] It can be understood that by introducing machine learning algorithms, especially neural network models with momentum terms, the microprocessor can more finely correct the results of preliminary estimates, improve measurement accuracy, and enhance the robustness of the model to abnormal situations.

[0037] Step S106, sending the final blood pressure value to the display module for output display.

[0038] Among them, the microprocessor sends the final blood pressure value to the display module, and the user can see the systolic and diastolic blood pressure values ​​clearly displayed on the screen, realizing an intuitive display of the blood pressure measurement results, which is convenient for users to understand their own health status in a timely manner; for example, "systolic blood pressure is 120mmHg, diastolic blood pressure is 80mmHg" is directly output on the display screen.

[0039] In the above implementation, from air pressure control to data analysis and then to output display, by combining the feature processing of oscillation waves and deep learning of neural networks, the efficiency and precision of the intelligent blood pressure measurement process are achieved, and the reliability and accuracy of blood pressure measurement results can still be ensured when facing complex physiological signals, thereby providing users with an efficient, convenient and reliable blood pressure monitoring solution, improving user experience, and providing more accurate and stable data support for the field of medical testing.

[0040] In addition, by introducing the momentum term parameter, the present application can achieve the smoothness and stability of the inflation and deflation process. Specifically, the momentum term parameter is similar to the proportional, integral and differential terms in the PID controller, which is used to adjust the working state of the air pump and the solenoid valve, thereby achieving a smoother pressure change. This method not only improves the controllability of the inflation and deflation process, but also reduces the measurement error caused by air pressure fluctuations, and significantly improves the accuracy and stability of blood pressure measurement. Ultimately, this control strategy provides a high-quality data basis for subsequent signal processing and blood pressure calculations, further enhancing the reliability and performance of the entire system.

[0041] Reference Figure 2 As an implementation method of step S103, the step of identifying the oscillation wave signal and extracting key features based on the eigenvalue method to obtain the oscillation wave characteristic parameters includes: Step S201, performing data preprocessing on the oscillation wave signal; Specifically, the data preprocessing step includes filtering and baseline correction to reduce the influence of external noise and baseline drift, and improve the accuracy of subsequent processing; In some embodiments, filtering can use a low-pass filter (such as a Butterworth filter) to remove high-frequency noise and retain useful low-frequency components; for example, a low-pass filter with a cutoff frequency of 5 Hz can be set to filter out noise above this frequency. Baseline correction can eliminate baseline drift caused by changes in cuff position or other factors by calculating the average value over a period of time and subtracting it from the signal; for example, the average value of the signal in the previous few seconds can be calculated as the baseline, and the entire signal can be subtracted from this baseline value.

[0042] Step S202, dividing the preprocessed oscillation wave signal into a plurality of sub-segments according to different preset pressure ranges; Specifically, the signal can be divided into multiple sub-segments according to a preset pressure range (such as one sub-segment for every 10 mmHg), each sub-segment represents a specific pressure interval, and the starting point and end point of each sub-segment are determined by detecting the peak point in each sub-segment. For example, during the deflation process, when the air pressure drops to the next preset pressure value, a new sub-segment begins.

[0043] It can be understood that, according to the different inflation and deflation stages, the signal is divided into multiple sub-segments to facilitate independent analysis in different pressure intervals, thereby improving the pertinence and accuracy of feature extraction.

[0044] Step S203, performing peak detection and trough detection on the oscillation wave signal in each sub-segment to determine the peak position and the trough position; Specifically, the peak detection can use the local maximum detection algorithm to perform peak detection on the oscillating wave signal in each sub-segment to find all obvious peak positions; the trough detection can further determine the trough position between adjacent peaks based on the peak position; step S204, calculate the characteristic parameters between each peak and trough to obtain the key characteristic parameters of the oscillating wave signal in each sub-segment; Among them, the key characteristic parameters of the oscillation wave signal in each sub-segment include but are not limited to the peak position, trough position, amplitude, frequency, period, rising edge slope and falling edge slope, etc.

[0045] Specifically, after determining the peaks and troughs, calculate the amplitude, frequency, period and other parameters between each peak and trough. For example, the amplitude can be calculated by the height difference between the peaks and the troughs, and the frequency can be estimated by the time interval between adjacent peaks. At the same time, identify and record other possible key features, such as the rising edge slope and falling edge slope of the waveform.

[0046] It can be understood that through detailed waveform feature extraction, the characteristics of the oscillation wave signal can be more comprehensively described, providing rich information for subsequent blood pressure calculation.

[0047] Step S205 , based on the principal component analysis method, dimension reduction processing and feature selection are performed on the key characteristic parameters of the oscillation wave signal in each sub-segment, and the selected features are normalized to obtain optimized oscillation wave characteristic parameters.

[0048] Among them, the principal component analysis (PCA) is applied to reduce the dimension of the feature parameters to extract the most representative principal components. For example, if there are 10 original features, they may be reduced to 3-5 main features through PCA. Then, according to the results of PCA, the most important features are selected; for example, the first three principal components with the largest contribution rate can be selected. Finally, the selected features are normalized to be within the same scale range for subsequent processing. For example, the minimum-maximum normalization method can be used to scale the feature values ​​to the interval [0, 1].

[0049] In the above implementation, the system can efficiently and accurately extract the key characteristic parameters of the oscillation wave signal. These steps not only improve the accuracy and stability of feature extraction, but also reduce data redundancy and enhance the robustness and scalability of the system.

[0050] Reference Figure 3 As an implementation of step S104, the step of calculating the preliminary blood pressure value based on the pre-established proportional coefficient model according to the air pressure data and the oscillation wave characteristic parameters includes: Step S301, loading a pre-established proportional coefficient model from system storage; Specifically, the proportional coefficient model is trained based on a large amount of clinical data and describes the relationship between systolic blood pressure, diastolic blood pressure and oscillation wave characteristic parameters.

[0051] It should be noted that when loading, the validity and version of the model are checked to ensure that the latest or specified version of the model is used.

[0052] Step S302, formatting the air pressure data and the oscillation wave characteristic parameters; Among them, the air pressure data and oscillation wave characteristic parameters are converted into a unified data format, such as converting all data into floating-point types to ensure the consistency and availability of input data, providing an accurate basis for subsequent calculations.

[0053] Step S303, mapping the oscillation wave characteristic parameters in each sub-segment to corresponding input variables in the proportional coefficient model to obtain a mapped characteristic parameter set; Among them, according to the requirements of the proportional coefficient model, the characteristic parameters of the oscillation wave in each sub-segment (such as amplitude, frequency, etc.) are mapped to the corresponding input variables in the model; for example, assuming that the input variables required by the model include amplitude, frequency and period, then these characteristic parameters in each sub-segment are respectively mapped to the input variables of the model.

[0054] Step S304, correcting the air pressure data, and time-aligning the corrected air pressure data with the characteristic parameter set to obtain an air pressure data set; Specifically, the air pressure data is corrected, for example, by smoothing uneven changes through interpolation or other methods; the corrected air pressure data is then time-aligned with the mapped feature parameter set to ensure that the data of the two at the same time point match.

[0055] Step S305, inputting the characteristic parameter set and the air pressure data set into the proportional coefficient model to calculate and obtain preliminary systolic pressure value and diastolic pressure value.

[0056] Specifically, the characteristic parameter set and the air pressure data set are substituted into the proportional coefficient model, and a preset proportional coefficient formula is applied for calculation.

[0057] In the above implementation, based on the pre-established proportional coefficient relationship, the air pressure data and the oscillation wave characteristic parameters are used to calculate the preliminary blood pressure value. This technical solution not only improves the accuracy and stability of the preliminary blood pressure estimation, but also provides a reliable data basis for subsequent neural network optimization, significantly improving the accuracy and reliability of the overall blood pressure measurement.

[0058] In one embodiment of the present application, the relationship of the proportionality coefficient model is: Systolic blood pressure = a1×amplitude+b1×frequency+c1×pressure+d1; Diastolic pressure = a2×amplitude+b2×frequency+c2×pressure+d2; In the above formula, a1, b1, c1, d1 and a2, b2, c2, d2 are predetermined proportional coefficients; since there is a linear relationship between systolic and diastolic blood pressure and the oscillation wave characteristic parameters, the above linear regression model can be used to fit this relationship.

[0059] Specifically, in order to establish the above-mentioned proportional coefficient model, a large amount of clinical data needs to be collected, which usually includes the following parts: blood pressure measurement sample data, such as systolic pressure and diastolic pressure; oscillation wave signal sample parameters, such as amplitude, frequency, period, rising edge slope and falling edge slope; air pressure sample data, that is, the real-time air pressure value in the cuff, recorded at different time points during inflation and deflation.

[0060] It should be noted that clinical data can be collected on a large scale in multiple medical institutions, and subjects of different ages, genders, and health conditions (such as hypertension, hypotension, and normotension) should be included as much as possible to cover a wide range of physiological conditions.

[0061] Reference Figure 4 A further embodiment of the present application further includes a training step of a blood pressure value prediction model, the training step including: Step S401, obtaining a sample data set and performing data preprocessing; The sample data set includes oscillation wave signal sample data, oscillation wave characteristic parameter sample data, preliminary blood pressure sample values ​​and corresponding blood pressure value label data, and the blood pressure value label data is the actually measured systolic and diastolic pressure; Specifically, the actually measured systolic and diastolic blood pressure values ​​are used as label data, and corresponded to each set of oscillation wave signal sample data, oscillation wave characteristic parameter sample data, and calculated preliminary blood pressure sample values, which constitute the data basis for training the model.

[0062] In some embodiments, the preprocessing steps include data cleaning, data normalization, data enhancement, etc., to ensure the consistency and availability of input data and provide a high-quality data foundation for subsequent training. Specifically, data cleaning is to remove outliers and noise to ensure the quality of the data. For example, statistical methods (such as Z-score) can be used to detect and remove outliers; data normalization refers to normalizing all feature parameters and blood pressure values ​​to the same scale range, such as [0,1] or [-1,1], which helps to improve the convergence speed and stability of the model; data enhancement such as random cropping, translation, scaling, etc., especially for oscillating wave signals, can enhance the diversity of data.

[0063] Step S402, dividing the sample data set into a training set, a validation set and a test set; The data set is divided into training set, validation set and test set. Usually 70% of the data is used as the training set, 15% as the validation set, and 15% as the test set. Random sampling can be used to divide the data to avoid data bias and ensure that the data distribution in each subset is consistent.

[0064] Step S403, inputting the training set into a pre-built convolutional neural network model for iterative training, calculating the loss function of the model, calculating the gradient through a back propagation algorithm, and optimizing the model parameters; Among them, the use of convolutional neural networks (CNN) to process time series data or multi-dimensional data is very effective. Through multi-layer convolution and pooling operations, features in the data are gradually extracted for decision-making in subsequent fully connected layers.

[0065] In one of the embodiments of the present application, the pre-constructed convolutional neural network model includes an input layer, a convolution layer, a fully connected layer, and an output layer; specifically, the input layer is set to receive the processed oscillation wave signal data and the preliminary blood pressure value, and splice them into a multidimensional array as input; then several convolution layers are constructed to extract high-order features in the oscillation wave signal, which can reveal subtle changes in the signal; the output of the convolution layer passes through several fully connected layers to further process and fuse the preliminary input blood pressure value; finally, the output layer is set with two neurons, one for predicting the systolic pressure value and the other for predicting the diastolic pressure value. Through multiple rounds of iterative training, the weights and biases of the model are gradually optimized, so that the model can accurately predict blood pressure values ​​from complex signals.

[0066] Step S404, updating the weights of the convolutional neural network model based on an optimization algorithm with a momentum term until the loss function meets a preset condition, thereby obtaining a trained convolutional neural network model; Specifically, when training the model, select an optimization algorithm with a momentum term, such as SGD with momentum. By introducing the momentum term, the gradient descent process is accelerated, the oscillation is reduced, and the convergence speed and stability of the model are improved. Then, the weights of the model are updated based on the calculated gradient and momentum terms. For example, when using SGD with momentum, the update formula is as follows: θ t =θ t-1 -v t In the above formula, v t is the speed (i.e. momentum) of the tth step, γ is the momentum coefficient (e.g. 0.9), η is the learning rate, is the loss function J at θt-1 The gradient of θ with respect to the parameter θ.

[0067] As an implementation of the loss function, the mean square error (MSE) can be used as the loss function. The loss function is used to measure the difference between the model predicted value and the true value. The gradient of the loss function is calculated through the back propagation algorithm, and the model parameters are adjusted using the optimization algorithm to minimize the loss function, so that the model can effectively learn and adjust its own parameters, thereby improving the accuracy and robustness of the prediction.

[0068] Step S405, verifying the convolutional neural network model based on the verification set, evaluating the performance of the convolutional neural network model and adjusting the hyperparameters of the model according to the verification results; The validation set data is input into the trained convolutional neural network to calculate the predicted systolic and diastolic blood pressures, and the evaluation indicators (such as MSE, MAE, R 2 The performance of the model is evaluated by using the validation set, and then the hyperparameters of the model (such as learning rate, batch size, momentum coefficient, etc.) are adjusted according to the performance evaluation results on the validation set to further optimize the model performance.

[0069] Step S406, testing the prediction ability of the adjusted convolutional neural network model based on the test set to obtain a blood pressure value prediction model.

[0070] Specifically, the test set data is input into the adjusted convolutional neural network to calculate the predicted systolic and diastolic blood pressure to evaluate its actual prediction ability and generalization performance. The performance indicators on the test set are the key criteria for measuring the practicality of the model. The mean square error (MSE), accuracy and other relevant evaluation indicators of the model can be evaluated to ensure that the model effect meets the practical standards.

[0071] In the above implementation, an optimization algorithm with a momentum term is used to iteratively train the convolutional neural network model, and the model performance is evaluated and optimized through a validation set and a test set. This technical solution not only improves the accuracy of blood pressure measurement, but also enhances the robustness of the model to abnormal situations. This optimization method significantly improves the accuracy and stability of blood pressure measurement, and provides a more reliable and efficient solution for the field of medical testing.

[0072] Reference Figure 5 As a further implementation of the intelligent blood pressure measurement method, after the step of sending the final blood pressure value to the display module for output display, it also includes: Step S501, determine whether the final blood pressure value exceeds the normal blood pressure range; if so, jump to step S502; if not, do not perform any operation.

[0073] In some embodiments, a normal blood pressure range may be defined according to medical standards; for example, the normal blood pressure range for adults is generally 90-120 mmHg for systolic pressure and 60-80 mmHg for diastolic pressure. By comparing the final blood pressure value with the defined normal blood pressure range, the user's blood pressure condition can be accurately determined and abnormal conditions can be discovered in a timely manner.

[0074] Step S502: Generate a blood pressure warning prompt based on the final blood pressure value and send it to the corresponding medical terminal.

[0075] If the final blood pressure value exceeds the normal range, a corresponding warning prompt is generated. In some embodiments, the warning prompt may include the following: the identity information of the subject (such as name, ID number, etc.), the final blood pressure value, the warning type (hypertension or hypotension), and the recommended measures (such as immediate medical treatment, close monitoring, etc.); the medical terminal may be a doctor's workstation, a nurse's handheld device, or other designated medical management system.

[0076] In the above implementation, after the final blood pressure value is sent to the display module for output display, it is further determined whether the blood pressure value exceeds the normal range. If it exceeds the normal range, a corresponding early warning prompt is generated and sent to the corresponding medical terminal. This technical solution can not only monitor the user's blood pressure status in real time, but also notify medical staff in time when abnormalities are found, thereby improving the response speed and quality of medical services. Through this early warning mechanism, patients can receive more timely medical intervention, reduce the risks caused by high blood pressure or low blood pressure, and improve the safety and effectiveness of overall medical care.

[0077] The embodiment of the present application also discloses an intelligent blood pressure measurement system based on the deflation oscillometric method.

[0078] An intelligent blood pressure measurement system based on deflation oscillometric method, the intelligent blood pressure measurement system comprising an air pump, a solenoid valve, a pressure sensor, a microprocessor, a display module and a cuff wrapped around the wrist or upper arm of a subject; An air pump, used to fill the cuff with gas; A solenoid valve, connected to the air pump, is used to control the inflow and outflow of gas in the cuff; A cuff for regulating internal air pressure via an air pump and a solenoid valve; A pressure sensor is connected to the cuff and is used to collect air pressure data and oscillation wave signals in the cuff in real time; A microprocessor is connected to the air pump, the solenoid valve and the pressure sensor, and is used to receive and process the air pressure data and the oscillation wave signal collected by the pressure sensor, and calculate the final blood pressure value; The display module is used to display the final blood pressure value calculated by the microprocessor.

[0079] Based on the above-mentioned intelligent blood pressure measurement system, in the actual operation process, first ensure that the hardware devices such as the air pump, solenoid valve, pressure sensor, etc. are correctly connected to the microprocessor, and the system is initialized; the subject starts the measurement program after wearing the cuff according to the instructions, and the microprocessor controls the air pump and solenoid valve to complete the inflation to block the blood flow and the subsequent deflation process; in this process, the pressure sensor collects air pressure and oscillation wave signals in real time, and the microprocessor pre-processes and extracts features of these signals; next, the proportional coefficient method is applied to calculate the preliminary blood pressure value, and then the preliminary result is optimized through the pre-trained neural network model to improve the measurement accuracy; finally, the final blood pressure value is displayed on the screen, and the data can be selectively stored or transmitted for further analysis and management; the system achieves efficient, accurate and reliable blood pressure measurement by combining traditional oscillometric method with advanced signal processing and machine learning technology.

[0080] As an implementation of a microprocessor, the microprocessor includes: An air pump control unit, used to control the air pump to inflate air into the cuff of the wrist or upper arm of the subject until a preset initial air pressure value is reached; A solenoid valve control unit, used for controlling the solenoid valve to release the gas in the cuff at a preset rate; A data receiving unit, used for receiving the air pressure data and oscillation wave signal in the cuff collected in real time by the pressure sensor; A feature extraction unit, used to identify the oscillation wave signal and extract key features based on the eigenvalue method to obtain oscillation wave characteristic parameters; A preliminary blood pressure value generating unit, used to calculate the preliminary blood pressure value according to the air pressure data and the oscillation wave characteristic parameters based on a pre-established proportional coefficient model; A final blood pressure value generating unit, used for inputting the oscillation wave signal, the oscillation wave characteristic parameter and the preliminary blood pressure value into a pre-trained blood pressure value prediction model with a momentum term to obtain a final blood pressure value; wherein the blood pressure value includes a systolic pressure value and a diastolic pressure value; The final blood pressure value sending unit is used to send the final blood pressure value to the display module for output display.

[0081] The intelligent blood pressure measurement system based on the deflation oscillometric method in the embodiment of the present application can implement any of the above-mentioned intelligent blood pressure measurement methods, and the specific working process of each module in the intelligent blood pressure measurement system can refer to the corresponding process in the above-mentioned method embodiment.

[0082] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are only illustrative; for example, the division of a certain module is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0083] The embodiment of the present application also discloses a computer device.

[0084] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above-mentioned intelligent blood pressure measurement method based on the deflation oscillometric method is implemented.

[0085] The embodiment of the present application also discloses a computer-readable storage medium.

[0086] A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any one of the above-mentioned intelligent blood pressure measurement methods based on the deflation oscillometric method.

[0087] Among them, computer-readable storage media can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus or device; the program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0088] It should be noted that in the above embodiments, the description of each embodiment has different emphases, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0089] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

Claims

1. An intelligent blood pressure measurement method based on deflation oscillometric method, applied to an intelligent blood pressure measurement system including an air pump, a solenoid valve, a pressure sensor, a microprocessor, a display module and a cuff, characterized in that: The measuring method comprises: Controlling the air pump to inflate air into the cuff on the wrist or upper arm of the subject until a preset initial air pressure value is reached; Controlling the solenoid valve to release the gas in the cuff at a preset rate, and receiving the air pressure data and oscillation wave signal in the cuff collected in real time by the pressure sensor; Identify the oscillation wave signal and extract key features based on the eigenvalue method to obtain oscillation wave characteristic parameters; Based on a pre-established proportional coefficient model, a preliminary blood pressure value is calculated according to the air pressure data and the oscillation wave characteristic parameters; Inputting the oscillation wave signal, the oscillation wave characteristic parameter and the preliminary blood pressure value into a pre-trained blood pressure value prediction model with a momentum term to obtain a final blood pressure value; wherein the blood pressure value includes a systolic pressure value and a diastolic pressure value; The final blood pressure value is sent to the display module for output display.

2. The intelligent blood pressure measurement method based on the deflation oscillometric method according to claim 1, characterized in that: The steps of identifying the oscillation wave signal and extracting key features based on the eigenvalue method to obtain oscillation wave characteristic parameters include: Performing data preprocessing on the oscillation wave signal; dividing the preprocessed oscillation wave signal into a plurality of sub-segments according to different preset pressure ranges; Performing peak detection and trough detection on the oscillation wave signal in each sub-segment, determining the peak position and the trough position, and calculating the characteristic parameters between each peak and trough to obtain the key characteristic parameters of the oscillation wave signal in each sub-segment; Based on the principal component analysis method, dimension reduction processing and feature selection are performed on the key characteristic parameters of the oscillation wave signal in each sub-segment, and the selected features are normalized to obtain the optimized characteristic parameters of the oscillation wave.

3. The intelligent blood pressure measurement method based on the deflation oscillometric method according to claim 2, characterized in that: Based on the pre-established proportional coefficient model, the step of calculating the preliminary blood pressure value according to the air pressure data and the oscillation wave characteristic parameters includes: Load a pre-built scale factor model from system storage; Formatting the air pressure data and oscillation wave characteristic parameters; Mapping the oscillation wave characteristic parameters in each sub-segment to corresponding input variables in the proportional coefficient model to obtain a mapped characteristic parameter set; Correcting the air pressure data, and time-aligning the corrected air pressure data with the characteristic parameter set to obtain an air pressure data set; The characteristic parameter set and the air pressure data set are input into a proportional coefficient model to calculate preliminary systolic pressure value and diastolic pressure value.

4. The intelligent blood pressure measurement method based on the deflation oscillometric method according to claim 3, characterized in that: The relationship of the proportional coefficient model is: Systolic blood pressure = a1×amplitude+b1×frequency+c1×pressure+d1; Diastolic pressure = a2×amplitude+b2×frequency+c2×pressure+d2; In the above formula, a1, b1, c1, d1 and a2, b2, c2, d2 are predetermined proportional coefficients.

5. The intelligent blood pressure measurement method based on deflation oscillometric method according to claim 1, characterized in that: The method further includes a training step of the blood pressure value prediction model, wherein the training step includes: Acquire a sample data set and perform data preprocessing; wherein the sample data set includes oscillation wave signal sample data, oscillation wave characteristic parameter sample data, preliminary blood pressure sample values ​​and corresponding blood pressure value label data, and the blood pressure value label data is the actually measured systolic and diastolic pressure; Dividing the sample data set into a training set, a validation set and a test set; Inputting the training set into a pre-built convolutional neural network model for iterative training, calculating the loss function of the model, calculating the gradient through a back-propagation algorithm, and optimizing the model parameters; The weights of the convolutional neural network model are updated based on an optimization algorithm with a momentum term until the loss function meets a preset condition, thereby obtaining a trained convolutional neural network model; Validating the convolutional neural network model based on the validation set, evaluating the performance of the convolutional neural network model and adjusting the hyperparameters of the model according to the validation results; The prediction ability of the adjusted convolutional neural network model is tested based on the test set to obtain the blood pressure value prediction model.

6. An intelligent blood pressure measurement method based on deflation oscillometric method according to any one of claims 1 to 5, characterized in that: After the step of sending the final blood pressure value to the display module for output display, the method further includes: Determining whether the final blood pressure value exceeds a normal blood pressure range; If so, a blood pressure warning prompt is generated according to the final blood pressure value and sent to the corresponding medical terminal.

7. An intelligent blood pressure measurement system based on deflation oscillometric method, characterized in that: The intelligent blood pressure measurement system includes an air pump, a solenoid valve, a pressure sensor, a microprocessor, a display module, and a cuff wrapped around the wrist or upper arm of the subject; The air pump is used to fill the cuff with gas; The solenoid valve is connected to the air pump and is used to control the inflow and outflow of gas in the cuff; The cuff is used to adjust the internal air pressure through the air pump and the solenoid valve; The pressure sensor is connected to the cuff and is used to collect air pressure data and oscillation wave signals in the cuff in real time; The microprocessor is connected to the air pump, the solenoid valve and the pressure sensor, and is used to receive and process the air pressure data and the oscillation wave signal collected by the pressure sensor, and calculate the final blood pressure value; The display module is used to display the final blood pressure value calculated by the microprocessor.

8. The intelligent blood pressure measurement system based on the deflation oscillometric method according to claim 7, characterized in that: The microprocessor comprises: An air pump control unit, used for controlling the air pump to inflate air into the cuff of the wrist or upper arm of the subject until a preset initial air pressure value is reached; A solenoid valve control unit, used for controlling the solenoid valve to release the gas in the cuff at a preset rate; A data receiving unit, used for receiving the air pressure data and oscillation wave signal in the cuff collected in real time by the pressure sensor; A feature extraction unit, used to identify the oscillation wave signal and extract key features based on a eigenvalue method to obtain oscillation wave feature parameters; a preliminary blood pressure value generating unit, configured to calculate the preliminary blood pressure value according to the air pressure data and the oscillation wave characteristic parameters based on a pre-established proportional coefficient model; A final blood pressure value generating unit, used for inputting the oscillation wave signal, the oscillation wave characteristic parameter and the preliminary blood pressure value into a pre-trained blood pressure value prediction model with a momentum term to obtain a final blood pressure value; wherein the blood pressure value includes a systolic pressure value and a diastolic pressure value; The final blood pressure value sending unit is used to send the final blood pressure value to the display module for output display.

9. A computer device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 6 when executing the program.

10. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and execute the method according to any one of claims 1 to 6.

Images