Blood Pressure Calibration Method, System, Device and Medium Based on Radial Basis Interpolation Model

Abstract

The present invention discloses a blood pressure calibration method based on a radial basis interpolation model, which specifically relates to the technical field of health data processing. The method includes: when the model blood pressure does not meet the accuracy requirements, determining whether the number of samples in the sample set is greater than M; if so, updating the sample set according to the first rule, and if not, updating the sample set according to the second rule; obtaining a target radial basis interpolation model based on the updated sample set to obtain the calibrated model blood pressure, so as to realize the calibration of the model blood pressure. After analyzing and comparing the model blood pressure to be calibrated and its corresponding calibrated blood pressure, the present invention calibrates the user's model blood pressure according to the radial basis interpolation model algorithm. As the number of calibration data provided by the user increases, an improved multi-method collaborative optimization method is used to optimize and improve the radial basis interpolation model. The present invention can quickly and effectively perform approximate correction for a small batch of samples, thereby improving the accuracy of the model blood pressure and enhancing the user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a blood pressure calibration method, system, device and apparatus based on a radial basis interpolation model. Background Art

[0002] At present, there are many software on the market for predicting and analyzing data such as blood pressure, heart rate, blood oxygen, etc. after collecting photoelectric and other signals by wearable devices. Among them, blood pressure is the result obtained by analyzing photoelectric signals. Due to the differences in individual physiological and psychological characteristics of users, the blood pressure model established using a unified model will have deviations among different individuals, and the deviation may be more than 20% at most, which brings doubts about the accuracy of the model to users, and thus the experience is very poor. In order to improve the accuracy of blood pressure measurement, some wearable devices have added a blood pressure calibration function item, but its accuracy still needs to be improved, and there is still a gap from the expectations of users.

[0003] Therefore, how to achieve personalized blood pressure calibration is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0004] In view of the defects in the prior art, the present invention provides a blood pressure calibration method, system, device and apparatus based on a radial basis interpolation model.

[0005] In a first aspect, a blood pressure calibration method based on a radial basis interpolation model includes:

[0006] When the model blood pressure does not meet the accuracy requirement, determine whether the number of samples in the sample set is greater than M;

[0007] If so, update the sample set according to the first rule, and if not, update the sample set according to the second rule;

[0008] Obtain a target radial basis interpolation model according to the updated sample set, and obtain the calibrated model blood pressure to achieve the calibration of the model blood pressure.

[0009] Preferably, before determining whether the number of samples in the sample set is greater than M when the model blood pressure accuracy does not meet the requirement, it further includes:

[0010] Obtain the model blood pressure and the calibrated blood pressure;

[0011] Compare the model blood pressure with the calibrated blood pressure to determine whether the model blood pressure meets the accuracy requirement.

[0012] Preferably, the updating the sample set according to the first rule includes:

[0013] If the distribution of the model blood pressures is dense, arrange the sample composed of the model blood pressure and the calibrated blood pressure behind the M samples;

[0014] If the model blood pressures are evenly distributed, they will be sorted according to the acquisition time series, and the sample composed of the earliest model blood pressure and the calibrated blood pressure will be placed at the end.

[0015] Preferably, the obtaining of the target radial basis interpolation model based on the updated sample set to obtain the calibrated model blood pressure includes:

[0016] Obtain a radial basis interpolation model;

[0017] Divide the sample set into a training set and a test set, optimize and train the radial basis interpolation model using the training set to obtain a target radial basis interpolation model, and verify the target radial basis interpolation model using the test set;

[0018] Calibrate the model blood pressure using the target radial basis interpolation model to obtain the calibrated model blood pressure.

[0019] Preferably, the updating of the sample set according to the second rule includes:

[0020] Directly add the sample formed by combining the model blood pressure and the calibrated blood pressure to the sample set.

[0021] Preferably, the obtaining of the target radial basis interpolation model based on the updated sample set to obtain the calibrated model blood pressure includes:

[0022] Obtain a radial basis interpolation model;

[0023] Radially optimize and train the radial basis interpolation model using the sample set to obtain a target radial basis interpolation model;

[0024] Calibrate the model blood pressure using the target radial basis interpolation model to obtain the calibrated model blood pressure.

[0025] In a second aspect, a blood pressure calibration system based on a radial basis interpolation model includes:

[0026] A judgment unit, configured to judge whether the number of samples in the sample set is greater than M when the model blood pressure does not meet the accuracy requirement;

[0027] A first update unit, configured to update the sample set according to the first rule if the number of samples in the sample set is greater than M, and update the sample set according to the second rule if the number of samples in the sample set is not greater than M;

[0028] A calibration unit, configured to obtain a target radial basis interpolation model based on the updated sample set to obtain the calibrated model blood pressure, so as to implement the calibration of the model blood pressure.

[0029] Preferably, it further includes:

[0030] A second update unit, configured to update the target radial basis interpolation model.

[0031] In a third aspect, a computer device includes a memory and a processor connected to the memory; the memory is used to store a computer program; the processor is used to run the computer program stored in the memory to execute any one of the above methods.

[0032] In a fourth aspect, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any one of the above methods can be implemented.

[0033] The beneficial effects of the present invention are reflected in:

[0034] The present invention provides a blood pressure calibration method based on a radial basis interpolation model. After analyzing and comparing the model blood pressure to be calibrated and its corresponding calibrated blood pressure, the user's model blood pressure is calibrated according to the radial basis interpolation model algorithm. As the calibration data provided by the user increases, an improved multi-method collaborative optimization method is used to optimize and improve the radial basis interpolation model, and the obtained target radial basis interpolation model is also better. The present invention can quickly and effectively perform approximate correction for small batches of samples, thereby improving the accuracy of the model blood pressure and enhancing the user experience. The present invention also provides a blood pressure calibration system, device, and medium based on a radial basis interpolation model. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0036] Figure 1 It is a flowchart of a blood pressure calibration method based on a radial basis interpolation model provided by the present invention;

[0037] Figure 2 It is a system block diagram of a blood pressure calibration system based on a radial basis interpolation model provided by the present invention;

[0038] Figure 3 It is a schematic structural diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0039] The following will describe in detail the embodiments of the technical solutions of the present invention with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, so they are only examples and cannot be used to limit the protection scope of the present invention.

[0040] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application shall have the ordinary meanings understood by those skilled in the art to which the present invention pertains.

[0041] Example 1

[0042] Reference Figure 1 , Figure 1 A blood pressure calibration method based on a radial basis interpolation model provided by an embodiment of the present invention includes:

[0043] Step 1: When the model blood pressure does not meet the accuracy requirement, determine whether the number of samples in the sample set is greater than M;

[0044] Specifically, before determining whether the number of samples in the sample set is greater than M when the model blood pressure accuracy does not meet the requirement, it further includes:

[0045] Obtain the model blood pressure and the calibrated blood pressure;

[0046] Compare the model blood pressure with the calibrated blood pressure to determine whether the model blood pressure meets the accuracy requirement.

[0047] It should be noted that the standard for determining whether it meets the accuracy requirement is to calculate the absolute value of the difference between the model blood pressure and the calibrated blood pressure. If the absolute value of the difference is less than the preset difference threshold, it meets the accuracy requirement. Exemplarily, the preset difference threshold is 5 mmHg. Of course, it can also be set according to the actual accuracy requirement and is not limited herein.

[0048] In the embodiment of the present invention, the model blood pressure is obtained by a rough measurement device such as a wearable device, and the calibrated blood pressure is obtained by a precise blood pressure collection device such as an electronic sphygmomanometer or a mercury sphygmomanometer. Exemplarily, the user measures blood pressure using a wearable device and records the wearable device data as follows: X 1 , X 2 ,…, Xn (referred to as the model blood pressure), and the calibration data Y 1 , Y 2 ,…, Yn (referred to as the calibrated blood pressure); generally, n = 3, not greater than 5.

[0049] Step 2: If so, update the sample set according to the first rule; if not, update the sample set according to the second rule;

[0050] Due to the limitations in solving the radial basis model, if the number of samples in the sample set is greater than M, and if the model blood pressure and calibrated blood pressure are directly added to the sample set as new samples, as the number of samples increases, the computational complexity of solving the model will also increase accordingly. At this time, we need to select the sample data and then re - model. Therefore, when the number of samples in the sample set is greater than M, the first rule is used to update the sample set, and the specific method is as follows: If the distribution of model blood pressures is dense, the samples composed of model blood pressure and calibrated blood pressure are arranged after the M samples; if the distribution of model blood pressures is uniform, they are sorted according to the acquisition time series, and the samples composed of the earliest model blood pressure and calibrated blood pressure are arranged at the end.

[0051] By updating the sample set through the above method, the distance relationship between model blood pressures and the acquisition time series relationship are considered, ensuring the uniform distribution of samples in the sample set, providing basic support for obtaining the target radial basis interpolation model subsequently, and improving the accuracy of the target radial basis interpolation model.

[0052] It should be understood that when the number of samples in the sample set is less than M, there is no need to process the original samples. Only the collected model blood pressure and calibrated blood pressure need to be added to the sample set as new samples to achieve the update of the sample set.

[0053] The samples composed of model blood pressure and calibrated blood pressure are directly added to the sample set.

[0054] Step three: According to the updated sample set, obtain the target radial basis interpolation model to get the calibrated model blood pressure, so as to achieve the calibration of the model blood pressure.

[0055] In an embodiment of the present invention, the obtaining the target radial basis interpolation model according to the updated sample set to get the calibrated model blood pressure includes:

[0056] Obtain the radial basis interpolation model;

[0057] Divide the sample set into a training set and a test set, optimize and train the radial basis interpolation model using the training set to obtain the target radial basis interpolation model, and verify the target radial basis interpolation model using the test set;

[0058] Calibrate the model blood pressure using the target radial basis interpolation model to get the calibrated model blood pressure.

[0059] Specifically, considering the relationship of the number of samples, in an embodiment of the present invention, M samples are used as the training set, and the remaining samples are used as the test set. The model is trained using the training set and verified using the test set. The multi - method collaborative optimization method is adopted, with the goal of minimizing the verification sample error, to obtain the target radial basis interpolation model for calibrating the blood pressure of individual users.

[0060] It should be noted that the Radial Basis Function (RBF) interpolation model uses a linear combination of a series of radial basis functions for approximate interpolation, and its expression is as follows:

[0061]

[0062] In the formula is the weight coefficient vector of each sample point, which is determined by the following formula:

[0063] β = A -1 y (2)

[0064] The matrix A is the basis function matrix, which is determined by the following formula:

[0065]

[0066] Wherein, is the radial basis function. Currently, commonly used basis functions include linear functions, cubic polynomial functions, thin plate spline functions, Gaussian functions, and MQ-type functions, etc. ||x - x i || represents the Euclidean distance r from a certain point x to the center point x i . Through r, the radial basis function can transform a multi-dimensional problem into a one-dimensional problem with r as the independent variable. Exemplarily, in the embodiment of the present invention, the basis function of the model selects a linear basis function: φ(r) = cr.

[0067] In another embodiment of the present invention, obtaining the target radial basis interpolation model according to the updated sample set and obtaining the calibrated model blood pressure includes:

[0068] Obtain the radial basis interpolation model;

[0069] Use the sample set to perform radial optimization training on the radial basis interpolation model to obtain the target radial basis interpolation model;

[0070] Use the target radial basis interpolation model to calibrate the model blood pressure to obtain the calibrated model blood pressure.

[0071] Due to the difference in the sample set, the obtained target radial basis interpolation models are different, and the calibration accuracies are also different. In actual operation, as the number of user model blood pressure collections increases and the target radial basis interpolation model is updated, the calibration accuracy will become higher and higher.

[0072] In summary, the present invention provides a blood pressure calibration method based on a radial basis interpolation model. After analyzing and comparing the model blood pressure to be calibrated and its corresponding calibrated blood pressure, the calibration of the user's model blood pressure is performed according to the radial basis interpolation model algorithm. As the number of calibration data provided by the user increases, an improved multi-method collaborative optimization method is used to optimize and improve the radial basis interpolation model, obtaining a target radial basis interpolation model, and the calibration of the model blood pressure is achieved through the target radial basis interpolation model. The present invention can quickly and effectively perform approximate correction for a small batch of samples, thereby improving the accuracy of the model blood pressure and enhancing the user experience.

[0073] Embodiment 2

[0074] Reference Figure 2 , Figure 2 A blood pressure calibration system based on a radial basis interpolation model provided by an embodiment of the present invention includes:

[0075] A judgment unit, configured to judge whether the number of samples in the sample set is greater than M when the model blood pressure does not meet the accuracy requirement;

[0076] A first update unit, configured to update the sample set according to a first rule if the number of samples in the sample set is greater than M, and update the sample set according to a second rule if the number of samples in the sample set is not greater than M;

[0077] A calibration unit, configured to obtain a target radial basis interpolation model according to the updated sample set, and obtain the calibrated model blood pressure to achieve the calibration of the model blood pressure.

[0078] In an embodiment of the present invention, the system further includes a second update unit for updating the target radial basis interpolation model. There is initially a basic radial basis interpolation model in the system. After obtaining the model blood pressure of a large number of users, a new radial basis interpolation model for calibration, that is, the target radial basis interpolation model, is established. The basic radial basis interpolation model is replaced with the target radial basis difference model, realizing the update of the target basis interpolation model to achieve the calibration of the model blood pressure for individual users.

[0079] The blood pressure calibration system based on a radial basis interpolation model provided by an embodiment of the present invention and the blood pressure calibration method based on a radial basis interpolation model provided by the foregoing embodiment are based on the same inventive concept. Therefore, for the more specific working processes of each unit in this embodiment, reference may be made to the corresponding content disclosed in the foregoing embodiment, and details are not described herein again.

[0080] Embodiment 3:

[0081] Reference Figure 3 , Figure 3A computer device provided by an embodiment of the present invention, the computer device includes a memory and a processor connected to the memory; the memory is used to store a computer program; the processor is used to run the computer program stored in the memory to execute any one of the above-mentioned blood pressure calibration methods based on a radial basis interpolation model. For the specific steps of the blood pressure calibration method based on a radial basis interpolation model, reference may be made to the corresponding content disclosed in the foregoing embodiments, and details will not be repeated herein.

[0082] It should be understood that in the embodiment of the present invention, the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0083] In the embodiment of the present invention, the memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A part of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0084] Embodiment Four

[0085] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it can implement any one of the above-mentioned blood pressure calibration methods based on a radial basis interpolation model. For the specific steps of the blood pressure calibration method based on a radial basis interpolation model, reference may be made to the corresponding content disclosed in the foregoing embodiments, and details will not be repeated herein.

[0086] The computer-readable storage medium may be the memory of the recognition device provided in the foregoing embodiments, such as the hard disk or memory of the recognition device. The computer-readable storage medium may also be an external memory of the device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the recognition device. Further, the computer-readable storage medium may further include both the internal memory and the external memory of the recognition device. The computer-readable storage medium is used to store the computer program and other programs and data. The computer-readable storage medium may also be used to temporarily store the data that has been output or is to be output.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features. 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 invention, and they should all be covered by the scope of the claims and the description of the present invention.

Claims

1. Blood pressure calibration method based on radial basis interpolation model, Characterized in that, Comprising: Obtaining model blood pressure and calibrated blood pressure; Comparing the model blood pressure with the calibrated blood pressure to determine whether the model blood pressure meets the accuracy requirements; When the model blood pressure does not meet the accuracy requirements, determining whether the number of samples in the sample set is greater than M; If so, updating the sample set according to the first rule, if not, updating the sample set according to the second rule; wherein, the updating the sample set according to the first rule includes: if the distribution of model blood pressures is dense, arranging the samples composed of model blood pressure and calibrated blood pressure after M samples; if the distribution of model blood pressures is uniform, sorting according to the acquisition time series and arranging the samples composed of the earliest model blood pressure and calibrated blood pressure at the end; According to the updated sample set, obtaining the target radial basis interpolation model to obtain the calibrated model blood pressure so as to realize the calibration of the model blood pressure.

2. The blood pressure calibration method based on radial basis interpolation model according to claim 1, Characterized in that, The obtaining the target radial basis interpolation model according to the updated sample set to obtain the calibrated model blood pressure includes: Establishing a radial basis interpolation model; Dividing the sample set into a training set and a test set, optimizing and training the radial basis interpolation model with the training set to obtain the target radial basis interpolation model, and verifying the target radial basis interpolation model with the test set; Calibrating the model blood pressure with the target radial basis interpolation model to obtain the calibrated model blood pressure.

3. The blood pressure calibration method based on radial basis interpolation model according to claim 1, Characterized in that, The updating the sample set according to the second rule includes: Directly adding the samples composed of model blood pressure and calibrated blood pressure to the sample set.

4. The blood pressure calibration method based on radial basis interpolation model according to claim 3, Characterized in that, The obtaining the target radial basis interpolation model according to the updated sample set to obtain the calibrated model blood pressure includes: Radially optimizing and training the radial basis interpolation model with the sample set to obtain the target radial basis interpolation model.

5. A blood pressure calibration system based on radial basis interpolation model, Characterized in that, Comprising: A judgment unit, configured to obtain model blood pressure and calibrated blood pressure; compare the model blood pressure with the calibrated blood pressure to determine whether the model blood pressure meets the accuracy requirements; when the model blood pressure does not meet the accuracy requirements, determine whether the number of samples in the sample set is greater than M; A first update unit, configured to update the sample set according to the first rule if the number of samples in the sample set is greater than M, and update the sample set according to the second rule if the number of samples in the sample set is not greater than M; wherein, the updating the sample set according to the first rule includes: if the distribution of model blood pressures is dense, arranging the samples composed of model blood pressure and calibrated blood pressure after M samples; if the distribution of model blood pressures is uniform, sorting according to the acquisition time series and arranging the samples composed of the earliest model blood pressure and calibrated blood pressure at the end; A calibration unit, configured to obtain the target radial basis interpolation model according to the updated sample set to obtain the calibrated model blood pressure so as to realize the calibration of the model blood pressure.

6. The blood pressure calibration system based on the radial basis interpolation model according to claim 5, wherein, it further comprises: a second update unit for updating the target radial basis interpolation model.

7. A computer device, wherein, the computer device includes a memory and a processor connected to the memory; the memory is used for storing a computer program; the processor is used for running the computer program stored in the memory to execute the method according to any one of claims 1-4.

8. A computer-readable storage medium, wherein, the storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1-4 can be implemented.

Images