A blood pressure measurement method, device and computer storage medium based on PPG technology
By using PPG technology in blood pressure measurement, multiple pulse wave characteristic quantities are extracted and pulse wave-blood pressure linear model is constructed, which solves the problems of low blood pressure measurement accuracy and poor individual adaptability in the prior art, and achieves higher measurement accuracy and adaptability.
Patent Information
- Application Number
- CN202311164611.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-09-11
- Publication Date
- 2025-01-03
- Estimated Expiration
- 2043-09-11
AI Technical Summary
The existing blood pressure measurement methods based on pulse wave characteristic parameters are not very accurate, and the fixed blood pressure model cannot adapt to the individual physiological characteristics of different subjects, resulting in inaccurate measurement results.
Using a blood pressure measurement method based on PPG technology, multiple pulse wave characteristic quantities, including Q value characteristics, use the LASSO regression algorithm to establish a pulse wave-blood pressure linear model, and provide general and private model selection , to improve measurement accuracy.
It improves the accuracy and adaptability of blood pressure measurement, and by constructing a personalized pulse wave-blood pressure linear model, it can more accurately represent the individual's blood pressure value, reducing measurement errors.
Smart Images
Figure CN117204835B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a blood pressure measurement method, device and computer storage medium, and in particular to a blood pressure measurement method, device and computer storage medium based on PPG technology. Background Art
[0002] Blood pressure is an important vital sign. Blood pressure measurement is extremely important for the health management and treatment of patients with hypertension and critical illnesses. The pulse wave characteristic parameter measurement method is a method of extracting characteristic information from the pulse waveform, finding its correlation with blood pressure, and then establishing a blood pressure model to measure blood pressure. This blood pressure measurement method based on pulse wave characteristic information does not require an inflatable cuff and can monitor the dynamic changes of blood pressure in real time.
[0003] The existing blood pressure model mainly extracts features such as pulse wave peak value, pulse wave trough value and pulse cycle from the pulse wave signal to fit the blood pressure to obtain the blood pressure model, and then uses the blood pressure model to calculate the corresponding diastolic pressure and systolic pressure after obtaining the pulse wave signal through re-testing. However, the test results of the blood pressure model obtained by this type of feature fitting are not very accurate.
[0004] In addition, existing blood pressure measurement devices usually use a fixed blood pressure model for measurement, and the physiological characteristics of the cardiovascular system of different subjects are very different. These inconsistencies in physiological characteristics have a profound impact on the collected pulse wave signal, and thus affect the blood pressure value. The blood pressure value predicted by a fixed blood pressure model cannot well represent an individual's blood pressure. Summary of the invention
[0005] In view of the above-mentioned defects of the prior art, the present invention provides a blood pressure measurement method based on PPG technology, aiming to improve the accuracy of blood pressure measurement based on pulse wave signals. The present invention also provides a blood pressure measurement device based on PPG technology and a computer storage medium.
[0006] The technical solution of the present invention is as follows: A blood pressure measurement method based on PPG technology comprises the following steps:
[0007] Step 1: collecting a pulse wave signal and a corresponding blood pressure value for a period of time, wherein the blood pressure value includes diastolic pressure and systolic pressure;
[0008] Step 2: pre-process the pulse wave signal and extract pulse wave feature quantities, wherein the pulse wave feature quantities are no less than three and include at least a Q value feature, and the Q value feature is calculated by the formula
[0009]
[0010] Determine, where the Q value corresponds to each pulse wave cycle, the Vs Corresponding to the maximum pulse wave amplitude, the V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle;
[0011] Step 3: establishing a pulse wave-blood pressure linear model through a regression algorithm based on the pulse wave characteristic and the blood pressure value;
[0012] Step 4: Collect the subject's pulse wave signal, extract the pulse wave feature of the subject's pulse wave signal, and obtain the subject's blood pressure value through the pulse wave-blood pressure linear model.
[0013] Furthermore, the pulse wave characteristic quantities include the peak-trough interval, the descending middle isthmus-trough interval, the descending middle wave-trough interval, the trough-trough interval, the descending middle isthmus relative height, the descending middle wave relative height and the Q value characteristic.
[0014] Furthermore, the regression algorithm is a LASSO regression algorithm.
[0015] Another object of the present invention is to improve the adaptability of the blood pressure measuring device to different subjects, so as to improve the accuracy of blood pressure measurement for each different subject. The corresponding technical solution is as follows.
[0016] A blood pressure measurement method based on PPG technology comprises the following steps:
[0017] Step 1, the subject selects the universal pulse wave-blood pressure linear model or the private pulse wave-blood pressure linear model measurement, and when the private pulse wave-blood pressure linear model measurement is selected, step 2 is entered, otherwise the subject's pulse wave signal is collected and the subject's blood pressure value is obtained through the universal pulse wave-blood pressure linear model;
[0018] Step 2: collecting a pulse wave signal and a corresponding blood pressure value for a period of time, wherein the blood pressure value includes diastolic pressure and systolic pressure;
[0019] Step 3: pre-process the pulse wave signal and extract pulse wave feature quantities, wherein the pulse wave feature quantities are no less than three and include at least a Q value feature, and the Q value feature is calculated by the formula
[0020]
[0021] Determine, where the Q value corresponds to each pulse wave cycle, the V s Corresponding to the maximum pulse wave amplitude, the V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle;
[0022] Step 4: establishing a pulse wave-blood pressure linear model through a regression algorithm based on the pulse wave characteristic and the blood pressure value;
[0023] Step 5: Collect the subject's pulse wave signal, extract the pulse wave feature of the subject's pulse wave signal, and obtain the subject's blood pressure value through the pulse wave-blood pressure linear model.
[0024] Furthermore, in step 2, the local device collects the pulse wave signal and the corresponding blood pressure value of a time period and uploads them to the server, the server executes steps 3 and 4 and sends the pulse wave-blood pressure linear model to the local device, and then the local device executes step 5.
[0025] The present invention also provides a blood pressure measurement device based on PPG technology, comprising:
[0026] An acquisition module, used for acquiring a pulse wave signal and a corresponding blood pressure value in a period of time, wherein the blood pressure value includes diastolic pressure and systolic pressure;
[0027] The feature extraction module preprocesses the pulse wave signal and extracts pulse wave feature quantities, wherein the pulse wave feature quantities are not less than three and include at least a Q value feature, and the Q value feature is calculated by the formula
[0028]
[0029] Determine, where the Q value corresponds to each pulse wave cycle, the V s Corresponding to the maximum pulse wave amplitude, the V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle;
[0030] A model building module, which builds a pulse wave-blood pressure linear model based on the pulse wave characteristic quantity and the blood pressure value through a regression algorithm;
[0031] And, a real-time measurement module collects the subject's pulse wave signal, extracts the pulse wave feature of the subject's pulse wave signal, and obtains the subject's blood pressure value through the pulse wave-blood pressure linear model.
[0032] Furthermore, it includes a selection module, in which the subject selects the general pulse wave-blood pressure linear model or the private pulse wave-blood pressure linear model for measurement. When the private pulse wave-blood pressure linear model is selected for measurement, the acquisition module, the feature extraction module, the model building module and the real-time measurement module are called to obtain the subject's blood pressure value; otherwise, the subject's pulse wave signal is collected and the subject's blood pressure value is obtained through the general pulse wave-blood pressure linear model.
[0033] The present invention also provides a computer storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned blood pressure measurement method based on PPG technology is implemented.
[0034] The advantages of the technical solution provided by the present invention are:
[0035] The present invention obtains Q value features as the main pulse wave feature quantity for constructing a pulse wave-blood pressure linear model by processing the pulse wave signal, and further constructs a pulse wave-blood pressure linear model with a LASSO regression algorithm, thereby improving the accuracy of blood pressure measurement based on the pulse wave. In addition, by setting a general pulse wave-blood pressure linear model and a private pulse wave-blood pressure linear model, and using personal data to update the private pulse wave-blood pressure linear model, more accurate blood pressure measurement can be provided for different subjects. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 Schematic diagram of a blood pressure measurement method based on PPG technology according to an embodiment. DETAILED DESCRIPTION
[0037] The present invention is further described below in conjunction with examples. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading this description, various equivalent modifications to this description by those skilled in the art fall within the scope defined by the claims attached to this application.
[0038] The embodiment of the present invention is a blood pressure measurement device based on PPG technology, including a selection module, a collection module, a feature extraction module, a model building module and a real-time measurement module.
[0039] The selection module is used to measure blood pressure using different detection models according to different choices of the subjects, that is, the subjects choose the universal pulse wave-blood pressure linear model or the private pulse wave-blood pressure linear model for measurement. The universal pulse wave-blood pressure linear model is a preset model, which is determined by pre-trained parameters and is suitable for most people, but it is less targeted. Since the pulse wave characteristics of different people are different, the deviation of blood pressure measurement results is slightly larger. The private pulse wave-blood pressure linear model can be adjusted according to the subject's situation to improve the accuracy of blood pressure measurement for different individuals.
[0040] When the private pulse wave-blood pressure linear model measurement is selected, the acquisition module, feature extraction module, model building module and real-time measurement module are respectively called to obtain the subject's blood pressure value. Otherwise, the subject's pulse wave signal is collected and the subject's blood pressure value is obtained through the general pulse wave-blood pressure linear model.
[0041] The acquisition module is used to collect the pulse wave signal and the corresponding blood pressure value for a period of time. The blood pressure value includes diastolic pressure and systolic pressure. The sphygmomanometer and the watch under test (the acquisition module of the blood pressure measurement device based on PPG technology, the real-time measurement module) are local devices, both worn on the left hand. The watch is worn on the wrist to collect the sensor probe data required by the model. The sphygmomanometer cuff is worn on the arm, and the data collection time is set. The watch data is collected first, and once the watch data collection is completed, the blood pressure test of the sphygmomanometer is immediately started. The two operations are continuous, and the test data of the sphygmomanometer and the pulse wave signal data of the watch are uploaded to the server.
[0042] The feature extraction module is set at the server end, and extracts the pulse wave feature after filtering and denoising the pulse wave signal. The pulse wave feature is not less than three, and these feature quantities include the peak-trough interval, the descending gorge-trough interval, the descending wave-trough interval, the trough-trough interval, the descending gorge relative height, the descending wave relative height and the Q value feature, wherein the Q value feature is obtained by the formula
[0043]
[0044] Determine, where the solution of Q value corresponds to each pulse wave cycle, V s Corresponding to the maximum pulse wave amplitude, V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle.
[0045] The model building module is set on the server side. It builds a pulse wave-blood pressure linear model based on the pulse wave feature and the blood pressure value through the LASSO regression algorithm. In this embodiment, the pulse wave-blood pressure linear model is sent to the watch as a private model.
[0046] The real-time measurement module is a local module, which is set in a watch in this embodiment. It collects the subject's pulse wave signal, extracts the pulse wave feature of the subject's pulse wave signal, and obtains the subject's blood pressure value through a pulse wave-blood pressure linear model.
[0047] Please combine Figure 1 As shown, the main process of the measurement method of the blood pressure measurement device based on the PPG technology is as follows:
[0048] Step 1: The subject selects the universal pulse wave-blood pressure linear model or the private pulse wave-blood pressure linear model measurement on the watch. When the private pulse wave-blood pressure linear model measurement is selected, step 2 is entered. Otherwise, the subject's pulse wave signal is collected and the subject's blood pressure value is obtained through the universal pulse wave-blood pressure linear model preset in the watch. The pulse wave-blood pressure linear model preset by the watch is directly measured using the universal pulse wave-blood pressure linear model or the updated private pulse wave-blood pressure linear model. After the pulse wave signal data is obtained through the watch, the pulse wave feature of the subject's pulse wave signal is extracted and the corresponding blood pressure value is output by the universal or updated private pulse wave-blood pressure linear model.
[0049] Step 2: Collect the pulse wave signal and the corresponding blood pressure value for a period of time. The blood pressure value includes diastolic pressure and systolic pressure. The pulse wave signal can be collected at a frequency of 100 Hz for 3-5 minutes, and then the blood pressure value is collected.
[0050] Step 3: pre-process the pulse wave signal and extract the pulse wave feature. According to the need to establish a private pulse wave-blood pressure linear model, the pulse wave feature is not less than three, selected from the peak-trough interval, the mid-fall-trough interval, the mid-fall-trough interval, the trough-trough interval, the mid-fall relative height, the mid-fall relative height and the Q value feature. However, the selected pulse wave feature needs to include the Q value feature, which is calculated by the formula
[0051]
[0052] Determine, where the solution of Q value corresponds to each pulse wave cycle, V s Corresponding to the maximum pulse wave amplitude, V d Corresponding to the minimum pulse wave amplitude, V m Corresponding to the average value of the pulse wave amplitude in a pulse cycle. The peak-trough interval, the descending gorge-trough interval, the descending wave-trough interval, the trough-trough interval, the descending gorge relative height, and the descending wave relative height can be performed according to the prior art such as the technical solution disclosed in the Chinese patent application number 201910887347.5.
[0053] Step 4: Based on the above pulse wave feature and blood pressure value, a pulse wave-blood pressure linear model is established through the LASSO regression algorithm, which can be simply represented as y=wx+b, where y is a two-dimensional matrix composed of diastolic pressure and systolic pressure, and x is the pulse wave feature selected for use. The parameters w and b are determined through the LASSO regression algorithm. After the parameters w and b are determined, the private pulse wave-blood pressure linear model is sent to the watch to update the private pulse wave-blood pressure linear model of the watch.
[0054] Step 5: The subject keeps still with the watch for 1 minute to collect the pulse wave signal, extracts the pulse wave feature of the subject's pulse wave signal, and inputs the updated private pulse wave-blood pressure linear model to obtain the subject's blood pressure value (including systolic pressure and diastolic pressure).
[0055] It should be noted that the specific methods of the above embodiments can form a computer program product. Therefore, the computer program product implemented in this application can be stored on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.).
[0056] Blood pressure was measured comparatively by 12 subjects. Each subject was measured with a control sphygmomanometer and a blood pressure measurement based on PPG technology. The results are as follows:
[0057]
[0058]
[0059] The accuracy of the general pulse wave-blood pressure linear model calculated through data analysis is as follows:
[0060] General Model Systolic blood pressure Diastolic blood pressure Mean Difference 4.1 mmHg 6.8 mmHg Standard Deviation 13 mmHg 6.5 mmHg
[0061] The accuracy of the private pulse wave-blood pressure linear model calculated through data analysis is as follows:
[0062]
[0063]
Claims
1. A blood pressure measurement method based on PPG technology, characterized in that: The following steps are involved: Step 1: collecting a pulse wave signal and a corresponding blood pressure value for a period of time, wherein the blood pressure value includes diastolic pressure and systolic pressure; Step 2: pre-process the pulse wave signal and extract pulse wave feature quantities, wherein the pulse wave feature quantities are no less than three and include at least a Q value feature, and the Q value feature is calculated by the formula Determine, where the Q value corresponds to each pulse wave cycle, the V s Corresponding to the maximum pulse wave amplitude, the V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle; Step 3: establishing a pulse wave-blood pressure linear model through a regression algorithm based on the pulse wave characteristic and the blood pressure value; Step 4: Collect the subject's pulse wave signal, extract the pulse wave feature of the subject's pulse wave signal, and obtain the subject's blood pressure value through the pulse wave-blood pressure linear model.
2. The blood pressure measurement method based on PPG technology according to claim 1, characterized in that: The pulse wave characteristic quantities include the peak-trough interval, the descending middle isthmus-trough interval, the descending middle wave-trough interval, the trough-trough interval, the descending middle isthmus relative height, the descending middle wave relative height and the Q value characteristic.
3. The blood pressure measurement method based on PPG technology according to claim 1, characterized in that: The regression algorithm is a LASSO regression algorithm.
4. A blood pressure measurement method based on PPG technology, characterized in that: The following steps are involved: Step 1, the subject selects the universal pulse wave-blood pressure linear model or the private pulse wave-blood pressure linear model measurement, and when the private pulse wave-blood pressure linear model measurement is selected, step 2 is entered, otherwise the subject's pulse wave signal is collected and the subject's blood pressure value is obtained through the universal pulse wave-blood pressure linear model; Step 2: collecting a pulse wave signal and a corresponding blood pressure value for a period of time, wherein the blood pressure value includes diastolic pressure and systolic pressure; Step 3: pre-process the pulse wave signal and extract pulse wave feature quantities, wherein the pulse wave feature quantities are no less than three and include at least a Q value feature, and the Q value feature is calculated by the formula Determine, where the Q value corresponds to each pulse wave cycle, the V s Corresponding to the maximum pulse wave amplitude, the V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle; Step 4: establishing a pulse wave-blood pressure linear model through a regression algorithm based on the pulse wave characteristic and the blood pressure value; Step 5: Collect the subject's pulse wave signal, extract the pulse wave feature of the subject's pulse wave signal, and obtain the subject's blood pressure value through the pulse wave-blood pressure linear model.
5. The blood pressure measurement method based on PPG technology according to claim 4, characterized in that: In step 2, the local device collects the pulse wave signal and the corresponding blood pressure value for a period of time and uploads them to the server. The server executes steps 3 and 4 and sends the pulse wave-blood pressure linear model to the local device, and then the local device executes step 5.
6. A blood pressure measurement device based on PPG technology, characterized in that: include: An acquisition module, used for acquiring a pulse wave signal and a corresponding blood pressure value in a period of time, wherein the blood pressure value includes diastolic pressure and systolic pressure; The feature extraction module preprocesses the pulse wave signal and extracts pulse wave feature quantities, wherein the pulse wave feature quantities are not less than three and include at least a Q value feature, and the Q value feature is calculated by the formula Determine, where the Q value corresponds to each pulse wave cycle, the V s Corresponding to the maximum pulse wave amplitude, the V d Corresponding to the minimum pulse wave amplitude, V m Corresponds to the average value of the pulse wave amplitude within one pulse cycle; A model building module, which builds a pulse wave-blood pressure linear model through a regression algorithm based on the pulse wave feature and the blood pressure value; And, a real-time measurement module collects the subject's pulse wave signal, extracts the pulse wave feature of the subject's pulse wave signal, and obtains the subject's blood pressure value through the pulse wave-blood pressure linear model.
7. The blood pressure measurement device based on PPG technology according to claim 6, characterized in that: The invention comprises a selection module, wherein the subject selects the general pulse wave-blood pressure linear model or the private pulse wave-blood pressure linear model for measurement. When the private pulse wave-blood pressure linear model is selected for measurement, the acquisition module, the feature extraction module, the model building module and the real-time measurement module are called to obtain the blood pressure value of the subject; otherwise, the subject's pulse wave signal is acquired and the blood pressure value of the subject is obtained through the general pulse wave-blood pressure linear model.
8. The blood pressure measurement device based on PPG technology according to claim 6 or 7, characterized in that: The pulse wave characteristic quantities include the peak-trough interval, the descending middle isthmus-trough interval, the descending middle wave-trough interval, the trough-trough interval, the descending middle isthmus relative height, the descending middle wave relative height and the Q value characteristic.
9. The blood pressure measurement device based on PPG technology according to claim 6 or 7, characterized in that: It is characterized in that The regression algorithm is a LASSO regression algorithm.
10. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the blood pressure measurement method based on PPG technology described in any one of claims 1 to 5 is implemented.