WO2022095796A1

WO2022095796A1 - Physiological characteristic signal processing method, electronic device, chip, and readable storage medium

Info

Publication number: WO2022095796A1
Application number: PCT/CN2021/127382
Authority: WO
Inventors: 李露平; 陈茂林
Original assignee: 华为技术有限公司
Priority date: 2020-11-05
Filing date: 2021-10-29
Publication date: 2022-05-12
Also published as: CN114530246A

Abstract

Embodiments of the present application relate to the field of electronic devices, and provide a physiological characteristic signal processing method. A physiological characteristic optimization signal is obtained by performing filtering processing on an acquired original physiological characteristic signal; peak points of the physiological characteristic optimization signal are extracted by using a preset peak extraction algorithm to obtain a first peak point set; then, secondary peak extraction and optimization processing is performed on the first peak point set on the basis of the original physiological characteristic signal to obtain a second peak point set, so as to obtain physiological characteristics of a target individual by performing analysis on the basis of the second peak point set and the physiological characteristic optimization signal. The embodiments of the present application further provide an electronic device, a chip, and a computer-readable storage medium. According to the present application, a secondary peak extraction mechanism is introduced, so that physiological characteristic data can be acquired when a user is active, and thus the use experience of the user is improved, and real-time monitoring of physiological characteristics of the user is implemented.

Description

Physiological feature signal processing method, electronic device, chip and readable storage medium

This application claims the priority of the Chinese patent application with the application number 202011225485.6 and the application name "Physiological Feature Signal Processing Method, Electronic Device, Chip and Readable Storage Medium", which was submitted to the China Patent Office on November 05, 2020, all of which are The contents are incorporated herein by reference.

technical field

The present application relates to the field of terminal technology, and in particular, to a physiological characteristic signal processing method, electronic device, chip, and computer-readable storage medium.

Background technique

As people pay more attention to their own health, electronic devices with physiological function measurement are more and more popular. For example, wearable devices can collect the human body's photoplethysmography (PPG). Measurement of signs such as blood oxygen, exercise, sleep, etc. In order to ensure the accuracy of each measurement, as shown in Figure 1, the signal collected by the wearable device will go through processing processes such as signal filtering, signal peaking, signal quality detection, and physical sign measurement, in which the signal quality can be evaluated. , once the signal quality is assessed as poor, the signal collected by the wearable device will not be used for subsequent physical measurements.

The functions based on wearable devices are becoming more and more complex, such as the detection and even prediction of heart problems such as atrial fibrillation and premature beats. These functions have stricter requirements on the quality of signals collected by wearable devices, and sometimes require users to maintain a certain period of time. The static state can be measured successfully. When the user is in an active state (for example, walking), the signal quality collected by the wearable device is often poor, and the requirements for the signal processing capabilities of software and hardware are relatively high, resulting in the measurement function of the wearable device cannot be used normally in most active states. . Since the interference generated by the user in the active state changes dynamically, the interference generated by the same action of different users is different, and even the interference generated by the same user repeating the same action may also be different. The interference in the active state does not Does not follow fixed change rules. In order to improve the signal quality, the existing methods are generally to improve the hardware such as sensors, or to optimize the algorithm in the signal filtering and signal peaking stages, but the anti-interference effect of these methods is not obvious.

SUMMARY OF THE INVENTION

In view of this, it is necessary to provide a physiological characteristic signal processing method, which can overcome the above-mentioned problems, and can collect data and perform physical sign measurement when the user is in an active state, so as to improve the user experience.

A first aspect of the embodiments of the present application discloses a physiological feature signal processing method, which includes: filtering a collected physiological feature signal of a target individual to obtain a physiological feature optimization signal; extracting the physiological feature by using a preset peak extraction algorithm optimizing the peak points in the signal, and constructing a first peak point set based on the extracted peak points; performing optimization processing on the first peak point set based on the physiological characteristic signal to obtain a second peak point set, wherein the optimization The processing includes one or more of processing of adding peak points, processing of deleting peak points, and processing of updating peak points; and optimizing the signal analysis based on the second peak point set and the physiological characteristics to obtain the physiological characteristics of the target individual feature.

By adopting the technical solution, the physiological characteristic data can be collected and processed when the user is in an active state, the measurement accuracy is high, the user experience is improved, and the user's physiological characteristics can be monitored in real time.

In a possible implementation manner, the using a preset peak extraction algorithm to extract the peak points in the physiological characteristic optimization signal, and constructing a first peak point set based on the extracted peak points, includes: using the preset peak points The peak extraction algorithm extracts the peak points in the physiological characteristic optimization signal, and constructs the first peak point set based on the extracted peak points; or extracts the trough in the physiological characteristic optimization signal by using the preset peak extraction algorithm points, and the first set of peak points is constructed based on the extracted trough points.

By adopting the technical solution, it can be realized that only the peak points or the trough points are used to construct the peak point set, thereby reducing the amount of signal processing operations.

In a possible implementation manner, the performing optimization processing on the first set of peak points based on the physiological characteristic signal includes: modeling the physiological characteristic signal to obtain a corresponding value of the physiological characteristic signal. Physiological characteristic waveform; splitting the physiological characteristic waveform into an up-slope waveform segment and a down-slope waveform segment, and selecting an up-slope waveform segment or a down-slope waveform segment as the target waveform; The peak point is marked on the target waveform; the target waveform is divided into a plurality of waveform windows, and the initial interference degree of each waveform window is calculated according to a preset interference degree calculation algorithm; The peak point of the window is optimized, and the preset interference degree calculation algorithm is used to recalculate the interference degree of the waveform window after the optimization process, until the interference degree of the waveform window reaches the minimum value, and the waveform is completed. window optimization processing; and aggregating peak points included in each of the waveform windows that have completed the optimization processing to obtain the second peak point set.

By adopting this technical solution, it is possible to realize the optimization processing of adding, deleting, modifying and modifying the first peak point set, so that the interference degree of each waveform window can be minimized, and finally a more accurate peak point set can be obtained.

In a possible implementation manner, after the selecting the up-slope waveform or the down-slope waveform as the target waveform, the method further includes: simplifying the curve segment in the target waveform to a straight line segment including only head and tail endpoints.

By adopting the technical solution, the curve segment in the target waveform can be simplified into a straight segment, and the calculation amount of the subsequent interference degree calculation can be reduced.

In a possible implementation manner, the preset interference degree calculation algorithm includes: calculating a slope distance between any two straight line segments marked with the peak point in the waveform window, and comparing the calculated slope distance Perform normalization processing; calculate the length ratio between any two straight line segments marked with the peak point in the waveform window, and perform normalization processing on the calculated length ratio; calculate any arbitrary length ratio in the waveform window. The absolute value of the lateral distance difference between the two straight line segments marked with the peak point, and the calculated absolute value of the lateral distance difference is normalized; The absolute value of the longitudinal distance difference between the straight line segments of the peak point, and the calculated absolute value of the longitudinal distance difference is normalized; and based on the normalization result of the slope distance, the length ratio From the normalization result, the normalization result of the absolute value of the horizontal distance difference, and the normalization result of the absolute value of the vertical distance difference, the interference degree of the waveform window is obtained.

By adopting this technical solution, the interference degree of the waveform window can be calculated based on the four dimensions of slope distance, length ratio, horizontal distance difference and vertical distance difference.

In a possible implementation manner, performing normalization processing on the calculated slope distances includes: performing normalization processing on the plurality of calculated slope distances respectively, and summarizing the normalization of each of the slope distances result; or accumulating multiple calculated slope distances to obtain a total slope distance, and performing normalization processing on the total slope distance.

By adopting this technical solution, the slope distance can be normalized to convert the slope distance into the interference degree.

In a possible implementation manner, the longer straight line segment of the two straight line segments is the denominator of the length ratio, and the normalizing the calculated length ratio includes: Normalize each length ratio separately, and summarize the normalization results of each length ratio; or accumulate multiple length ratios calculated to obtain a total length ratio, and normalize the total length ratio processing.

By adopting this technical solution, the length ratio can be normalized to convert the length ratio into the interference degree.

In a possible implementation manner, the normalizing the calculated absolute values of the lateral distance differences includes: performing an average operation on the absolute values of multiple calculated lateral distance differences to obtain an average lateral distance difference. distance difference; and normalizing the calculated absolute value of each lateral distance difference based on the average lateral distance difference, and summarizing the normalized result of the absolute value of each lateral distance difference.

By adopting this technical solution, the absolute value of the lateral distance difference can be normalized, so as to convert the absolute value of the lateral distance difference into the degree of interference.

In a possible implementation manner, performing normalization processing on the calculated absolute values of longitudinal distance differences includes: performing an average operation on the absolute values of multiple calculated longitudinal distance differences to obtain an average longitudinal distance difference. distance difference; and normalizing the calculated absolute value of each longitudinal distance difference separately based on the average longitudinal distance difference, and summarizing the normalized result of the absolute value of each longitudinal distance difference.

By adopting this technical solution, the absolute value of the longitudinal distance difference can be normalized, so as to convert the absolute value of the longitudinal distance difference into the degree of interference.

In a possible implementation manner, the performing optimization processing on the peak point of the waveform window includes: searching for abnormal changes in slope distance, abnormal length ratio changes, abnormal changes in horizontal distance difference, or abnormal changes in vertical distance in the waveform window. The area where the distance difference changes abnormally, and the peak point in the area is optimized.

By adopting this technical solution, it is possible to quickly locate areas that may require peak point optimization processing, saving optimization time.

In a possible implementation manner, the performing optimization processing on the peak point of the waveform window includes: when the initial interference degree of the waveform window is greater than or equal to a preset interference degree, performing optimization processing on the peak value of the waveform window. Click to optimize.

By adopting this technical solution, an optimization attempt can be made for a waveform window with room for improvement in signal quality, and optimization efficiency can be improved.

In a possible implementation manner, the method further includes: when the initial interference degree of the waveform window is less than the preset interference degree, abandoning the optimization process for the peak point of the waveform window.

By adopting this technical solution, it is possible to avoid ineffective optimization attempts for waveform windows with poor signal quality, and to save optimization time.

In a possible implementation manner, the optimizing the signal analysis based on the second peak point set and the physiological characteristics to obtain the physiological characteristics of the target individual includes: analyzing the second peak point set and the physiological characteristics Perform signal quality evaluation on the characteristic optimization signal; and when the signal quality evaluation result is good signal quality, analyze and obtain the physiological characteristic of the target individual based on the second peak point set and the physiological characteristic optimization signal.

By adopting the technical solution, the signal quality can be evaluated, and only the signal evaluated as the signal quality number will be used for the subsequent physical measurement.

In a second aspect, an embodiment of the present application provides a computer-readable storage medium, including computer instructions, which, when the computer instructions are executed on an electronic device, cause the electronic device to execute the physiological characteristic signal processing method described in the first aspect.

In a third aspect, an embodiment of the present application provides an electronic device, the electronic device includes a processor and a memory, the memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the electronic device The physiological characteristic signal processing method according to the first aspect is performed.

In a fourth aspect, an embodiment of the present application provides a computer program product, which, when the computer program product runs on a computer, causes the computer to execute the physiological characteristic signal processing method described in the first aspect.

In a fifth aspect, an embodiment of the present application provides an apparatus, and the apparatus has a function of implementing the behavior of the electronic device in the method provided in the first aspect. The functions can be implemented by hardware, or by executing corresponding software by hardware. The hardware or software includes one or more modules corresponding to the above functions.

It can be understood that the computer-readable storage medium described in the second aspect, the electronic device described in the third aspect, the computer program product described in the fourth aspect, and the apparatus described in the fifth aspect are all the same as the above-mentioned first aspect. The method of the aspect corresponds to, therefore, the beneficial effects that can be achieved can be referred to the beneficial effects of the corresponding methods provided above, which will not be repeated here.

Description of drawings

FIG. 1 is a schematic flow chart of an existing wearable device performing physiological feature signal processing;

FIG. 2 is a schematic flowchart of a physiological characteristic signal processing method provided by an embodiment of the present application;

3 is a schematic waveform diagram of a segment of a PPG signal detected by an electronic device provided by an embodiment of the present application;

4 is a schematic diagram of a waveform in which the PPG signal of FIG. 3 only retains a down-slope waveform segment and is marked with a first set of peak points;

5 is a waveform schematic diagram in which the curve segment in the downslope waveform segment of FIG. 4 is simplified into a straight line segment including only the head and tail endpoints;

6 is a schematic flowchart of a physiological characteristic signal processing method provided by another embodiment of the present application;

FIG. 7 is a schematic structural diagram of a possible electronic device provided by an embodiment of the present application.

Detailed ways

It should be noted that, in this application, "at least one" refers to one or more, and "a plurality" refers to two or more. "And/or", which describes the relationship between the associated objects, means that there can be three relationships, for example, A and/or B can mean: A alone exists, A and B exist at the same time, and B exists alone, where A, B Can be singular or plural. The terms "first", "second", "third", "fourth", etc. (if present) in the description and claims of this application and the drawings are used to distinguish similar objects, not to Describe a particular order or sequence.

For ease of understanding, some descriptions of concepts related to the embodiments of the present application are exemplarily given for reference.

Referring to FIG. 2 , a physiological feature signal processing method provided by an embodiment of the present application is applied to an electronic device 100 , and the electronic device 100 may be a smart watch, a smart bracelet, a physical sign measuring instrument, etc., and has a physiological feature measuring function device of. In this embodiment, the physiological characteristic signal processing method may include:

21. Perform filtering processing on the collected physiological characteristic signal of the target individual to obtain a physiological characteristic optimization signal.

In some embodiments, by using a preset filtering method to filter the collected physiological characteristic signal, some noises contained in the physiological characteristic signal can be filtered out to obtain an optimized physiological characteristic signal. The preset filtering method can select an existing sign signal filtering algorithm according to actual needs, such as a wavelet decomposition algorithm, a frequency domain analysis algorithm, a modal decomposition algorithm (such as empirical mode decomposition), an independent component analysis algorithm, Adaptive filtering algorithms, etc.

For example, the physiological characteristic signal is a PPG signal, and the target individual is the wearer of the electronic device 100 . The filtered and optimized PPG signal can be obtained by filtering the PPG signal of the target individual collected by the electronic device 100 .

22. Extract the peak points in the physiological characteristic optimization signal, and construct a first peak point set based on the extracted peak points.

In some embodiments, a preset peak-lifting algorithm may also be used to extract the peak points in the physiological characteristic optimization signal, the peak points may be peak points or trough points, and the extracted peak points may be grouped into a set, Then, the first peak point set is obtained. That is, a preset peak-lifting algorithm may be used to extract the peak points in the physiological characteristic optimization signal, and a first set of peak points may be constructed based on the extracted peak points, or a preset peak-lifting algorithm may be used to extract the peak points in the physiological characteristic optimization signal. , and construct the first set of peak points based on the extracted trough points. The following takes the peak point as the peak point as an example for illustration.

The preset peak-lifting algorithm may also select an existing physical-signal signal peak-lifting algorithm according to actual needs, such as a Bayesian decision classification algorithm, a machine learning classification algorithm, a heuristic algorithm, and the like.

23. Perform optimization processing on the first peak point set based on the physiological characteristic signal to obtain a second peak point set.

In some embodiments, the optimization process may include one or a combination of processing of adding a peak point, deleting a peak point, and updating a peak point. The processing of adding a peak point may refer to adding a peak point to the first peak point set, and the processing of deleting a peak point may refer to deleting a peak point from the first peak point set, and the updating The peak point processing may refer to deleting a peak point from the first peak point set and adding a peak point to the first peak point set at the same time.

In some embodiments, the physiological characteristic signal is an example of a PPG signal for illustration. The PPG signal collected by the electronic device 100 can be modeled, and the PPG signal can be converted into the physiological characteristic waveform S1 shown in FIG. The light intensity detected by the optical heart rate sensor is the dimension), and then the physiological characteristic waveform is divided into an up-slope waveform segment S11 and a down-slope waveform segment S12, and then one waveform segment can be arbitrarily selected as the target waveform for subsequent analysis. The following is an example of selecting the down-slope waveform segment S12 as the target waveform.

Select the down-slope waveform segment S12 in the physiological characteristic waveform S1 shown in FIG. 3 to obtain a waveform diagram as shown in FIG. 4 . Further, each peak point in the first peak point set may be marked on the target waveform shown in FIG. 4 . Since the target waveform contains many waveform segments, in order to speed up the signal analysis, the target waveform can be divided into multiple waveform windows, and then the interference degree of each waveform window can be calculated by using the preset interference degree calculation algorithm.

In some embodiments, the target waveform may be divided into multiple waveform windows with time as the segmentation dimension, and each waveform window contains the same time scale, for example, each waveform window contains a 1-second downslope waveform segment. The target waveform can also be divided into multiple waveform windows with the number of waveform segments as the segmentation dimension, and each waveform window contains the same number of waveform segments, for example, each waveform window contains 30 downslope waveform segments.

In some embodiments, in order to reduce the amount of calculation of the interference degree, the curve segment in each waveform window may be simplified to a straight line segment including only the head and tail endpoints, and then the interference degree calculation is performed. As shown in FIG. 5 , the curve segment shown in FIG. 4 is simplified to a straight line segment including only the head and tail end points. Using a preset interference degree calculation algorithm to calculate the interference degree of the waveform window may include: a. Calculate the slope distance between any two straight line segments marked with the peak point in the waveform window (that is, any two straight line segments marked with the peak point) The included angle between the straight line segments), and normalize the calculated slope distance; b. Calculate the length ratio between any two straight line segments marked with the peak point in the waveform window (the longer one The length value of the straight line segment is the denominator of the length ratio), and normalize the calculated length ratio; c. Calculate the absolute value of the lateral distance difference between any two straight line segments marked with peak points in the waveform window , and normalize the absolute value of the calculated horizontal distance difference; d. Calculate the absolute value of the vertical distance difference between any two straight line segments marked with the peak point in the waveform window, and compare the calculated The absolute value of the vertical distance difference is normalized; e. based on the normalized result of the slope distance, the normalized result of the length ratio, the normalized result of the absolute value of the horizontal distance difference and The normalized result of the absolute value of the longitudinal distance difference is calculated to obtain the interference degree of the waveform window.

In some embodiments, the larger the value of the slope distance, the larger the result of the normalization process. Performing normalization processing on the calculated slope distances may refer to performing normalization processing on multiple calculated slope distances respectively (the closer the value of the slope distances is to 0°, the smaller the result obtained by performing the normalization processing, the smaller the slope distance is. The closer the value of the distance is to 90°, the larger the result obtained from the normalization process), convert to obtain the corresponding degree of interference, and then summarize the normalized results, which can also mean that the calculated slope distances are accumulated first. The total slope distance is obtained, and then the total slope distance is normalized to obtain the corresponding interference degree.

In some embodiments, the smaller the value of the length ratio, the larger the result of the normalization process. Performing normalization processing on the calculated length ratios may refer to performing normalization processing on multiple calculated length ratios respectively (the closer the value of the length ratios is to 0, the larger the result obtained by performing the normalization processing, and the longer the length ratios are. The closer the value is to 1, the smaller the result obtained by normalization), convert to obtain the corresponding degree of interference, and then summarize the normalized results, which can also mean that the calculated slope distances are first accumulated to obtain the total The slope distance is then normalized to the total slope distance, and the corresponding interference degree is obtained by conversion.

In some embodiments, normalizing the calculated absolute value of the lateral distance difference may include: firstly averaging the absolute values of multiple calculated lateral distance differences to obtain the average lateral distance difference of the waveform window , and then normalize the absolute values of the multiple calculated lateral distance differences (the closer the absolute value of the lateral distance difference is to the average lateral distance difference, the smaller the result obtained by normalization, and the absolute value of the lateral distance difference is smaller. The more the value deviates from the average lateral distance difference, the larger the result obtained by normalization), the corresponding interference degree is obtained by conversion, and then the normalization results are summarized.

In some embodiments, normalizing the calculated absolute value of the longitudinal distance difference may include: first averaging the absolute values of the plurality of calculated longitudinal distance differences to obtain the average longitudinal distance difference of the waveform window , and then normalize the calculated absolute values of the multiple longitudinal distance differences respectively (the closer the absolute value of the longitudinal distance difference is to the average longitudinal distance difference, the smaller the result obtained by normalization, and the absolute value of the longitudinal distance difference is smaller. The more the value deviates from the average longitudinal distance difference, the larger the result obtained by normalization), the corresponding interference degree is obtained by conversion, and then the normalization results are summarized.

In the process of calculating the interference degree of the waveform window, the interference degree of the waveform window is divided into four dimensions: slope distance, length ratio, horizontal distance difference, and vertical distance difference for calculation and normalization. Interference degree, and then accumulate the normalized results of the slope distance, the normalized results of the length ratio, the normalized results of the absolute value of the horizontal distance difference, and the normalized results of the absolute value of the vertical distance difference, you can get The noise level of this waveform window.

In some embodiments, when the initial interference degree of each waveform window is obtained by calculation, it may be determined whether the initial interference degree is greater than a preset interference degree to determine whether it is necessary to adjust the peak point in the waveform window. When the initial interference degree of the waveform window is greater than the preset interference degree, it indicates that the quality of the peak points in the waveform window is poor, and even through subsequent optimization processing, the existing signal quality cannot be verified. When the initial interference degree of the waveform window is greater than or less than the preset interference degree, it indicates that the quality of the peak points in the waveform window has room for adjustment, and the existing signal quality check may be passed through the subsequent optimization processing of the present application. The size of the preset interference degree can be set according to actual needs.

In some embodiments, when trying to optimize the peak point of the waveform window, the preset interference degree calculation algorithm may be used to recalculate the interference degree of the optimized waveform window. Repeat the calculation of the interference degree until the interference degree of the waveform window reaches the minimum value, and stop optimizing the waveform window. After the optimization process is completed for each waveform window, the peak points currently included in each waveform window for which the optimization process has been completed can be aggregated to construct a second peak point set.

As shown in Figure 5, an attempt to optimize the peak point of the waveform window can be to add a peak point to a straight line segment (the straight line segment was not marked with a peak point before), or delete the peak point marked by a straight line segment. , or delete the peak point marked by a straight line segment, and then add a new peak point on another straight line segment (the waveform segment was not marked with a peak point before).

In some embodiments, since the change of human body characteristics generally follows an approximate linear change principle, the sudden change generally does not have a large span. For example, to analyze the slope distance, length ratio, horizontal distance difference, and vertical distance difference of several adjacent straight line segments marked with peak points, if it is found that a certain value suddenly changes greatly, it may be necessary to analyze the peak point in this area. Perform optimization processing, try to add peak points and/or delete peak points, and recalculate the interference degree to try to minimize the interference degree of the waveform window and save the optimization processing time.

In some embodiments, the number of attempts of the optimization process may also be limited, and after completing the preset number of optimization processes, a waveform state with a minimum interference degree is selected.

In some embodiments, a polling adjustment method can also be used to try to add and/or delete peak points for each straight line segment in the waveform window, and recalculate the interference degree of the waveform window until the interference degree of the waveform window is reached. To obtain the minimum value, the processing time is relatively long compared to the optimized processing method described in the previous description.

24. Obtain the physiological characteristic of the target individual based on the second peak point set and the physiological characteristic optimization signal analysis.

In some embodiments, when the second peak point set is obtained, an existing physical sign measurement and analysis method (such as the signal quality detection and physical sign measurement steps in FIG. 1 ) can be used to optimize the signal for the second peak point set and physiological characteristics The analysis is carried out to obtain the physiological characteristics of the target individual. Such as physiological characteristics such as heart rate, blood pressure and so on.

For example, for heart rate measurement, the heart rate value is obtained by conversion based on the number of peak points within a certain period of time. For example, if the number of peak points of the 5s physiological characteristic optimization signal is N, then the heart rate is N*12.

In some embodiments, signal quality evaluation may be performed on the second peak point set and the physiological characteristic optimization signal. If the evaluation result is poor signal quality, this segment of signal will not be used for subsequent physical measurement, and the signal will be terminated directly. deal with. If the evaluation result is that the signal quality is good, this segment of the signal will be used for physical sign measurement, and the existing sign analysis method can be used to analyze the second peak point set and the physiological characteristic optimization signal to obtain the physiological characteristics of the target individual.

The above physiological characteristic signal processing method first uses the existing filtering and peak-lifting technology to obtain the initial peak point set, and then uses the secondary peak-lifting mechanism to optimize the initial peak point set to obtain the final peak point set, which can be realized when the user is in the Collecting data in an active state eliminates the need to deliberately keep the user in a stationary state for a long time, increases the applicable scenarios for physical sign measurement, improves the user experience, and truly realizes real-time monitoring of the user's physiological characteristics, which can improve application scenarios such as wearable devices.

Referring to FIG. 6 , a schematic flowchart of an electronic device 100 implementing physiological feature measurement for a target individual provided by an embodiment of the present application.

61. Signal collection of wearable devices. The target individual wears the electronic device 100, and the electronic device 100 can collect the original physiological characteristic signal.

62. Signal filtering processing. An existing filtering algorithm can be used to filter the original physiological characteristic signal to obtain an optimized physiological characteristic signal.

63. Signal peak processing. An existing peak-lifting algorithm may be used to perform a peak point extraction operation on the obtained physiological characteristic optimized signal obtained by filtering, so as to obtain a first peak point set. The way to lift the peaks can be to extract only the peak points or only the trough points.

64. Secondary peak-lifting treatment. Model the original physiological characteristic signal, and divide the waveform obtained by modeling into an up-slope waveform segment and a down-slope waveform segment, and then arbitrarily select a waveform segment as the target waveform for subsequent analysis, and then collect the first peak points. The included peak points are marked on the target waveform and the interference level is calculated. The interference degree is minimized by trying to add peak points, delete peak points, and update peak points on the target waveform, and then construct a second peak point set based on the peak points in the state of minimum interference degree.

65. Signal quality detection. Use the existing signal quality detection method to evaluate the second peak point set and the optimized signal of physiological characteristics. If the evaluation result is that the signal quality is poor, the signal will not be used for the subsequent physical measurement, and the signal processing will be ended directly. .

66. Physical sign measurement. When using the existing signal quality detection method to evaluate the second peak point set and the optimized signal of physiological characteristics, if the evaluation result is that the signal quality is good, this segment of signal will be used to measure the signs, and the existing signs can be used. The analysis method analyzes the second peak point set and the physiological characteristic optimization signal to obtain the physiological characteristics of the target individual.

The following is based on the experimental data of physiological feature measurement using the existing physiological feature signal processing method (the method shown in FIG. 1 ) and the experimental data of physiological feature measurement using the physiological feature signal processing method shown in FIG. 6 of the present application. Comparison instructions.

In the prior art, in order to ensure the accuracy of atrial fibrillation detection, the screening of signal quality is relatively strict, and many signals collected in active states cannot pass the verification of signal quality. Therefore, the atrial fibrillation detection function generally requires the user to remain still for about 1 minute before Perform atrial fibrillation testing.

experiment one:

In this experiment, it is no longer mandatory to keep the collected user in a still state, so some signal data of the user in the active state (such as walking) may be collected. Sample 1: 9899 segments of PPG signals without atrial fibrillation were collected in static and active states, and each segment lasted about 1 minute; sample 2: 6740 segments of PPG signals with atrial fibrillation were collected in static and active states, each segment The signal lasts about 1 minute.

The sample 1 and sample 2 are processed by the existing physiological characteristic signal processing method. Among the 9899 non-AF PPG signals collected, 6891 signals have passed the existing signal quality inspection, and it can be considered that most of them are in a static state. The remaining 3008 segments of the signal did not pass the existing signal quality check, and it can be considered that most of them were collected in the active state; among the 6740 segments of atrial fibrillation-free PPG signals collected, 3031 segments passed the existing signals. For quality inspection, it can be considered that most of the signals were collected in a static state, and the remaining 3709 segments of signals did not pass the existing signal quality inspection, and it can be considered that most of them were collected in an active state. That is, when using the existing physiological characteristic signal processing methods, about 6717 (3008+3709) segments of PPG signal data cannot pass the signal quality verification. In the data that passes the signal quality verification, the accuracy of atrial fibrillation measurement is above 95%. .

When using the physiological characteristic signal processing method shown in FIG. 6 of the present application to process sample 1 and sample 2, 1933 signals passed the signal quality inspection among the 3008 signals that did not pass the signal quality inspection before, and only 1075 signals remained. The segment signal did not pass the signal quality inspection. Among the 3709 segment signals that did not pass the signal quality inspection, 1995 segment signals passed the signal quality inspection, and only 1714 segment signals did not pass the signal quality inspection, that is, the PPG signals that failed to pass the signal quality verification. The number was reduced from 6717 segments to 2789 segments (1075+1714), more than half of the signal data collected in the active state was recalled, and the accuracy of atrial fibrillation measurement remained above 95%.

Experiment 2:

Provide atrial fibrillation warning to the user in the vehicle (driving or riding) state. The current atrial fibrillation warning function generally cannot be used when the user is in the vehicle. In this experiment, the PPG signal data in the vehicle scene was collected. In order to ensure the safety of the patient, the vehicle-mounted data of the atrial fibrillation patient was not collected, so the actual samples were negative Samples to test the false alarm rate of atrial fibrillation (the bumps in the vehicle state will cause the signal to fluctuate, making it easy to produce false alarms in the results measured by the current electronic equipment). Sample 1: 343-segment PPGs are collected from the driver's left hand in a vehicle-mounted scenario; Sample 2: 343-segment PPGs are collected from the driver's right hand in an on-board scenario; Sample 3: 49-segment PPGs are collected from the driver in a stationary scenario. That is, a total of 735 (343+343+49) segments of PPG are collected, and each segment of the signal lasts about 1 minute.

Among the collected 735-segment non-AF PPG signals, it can be considered that most of them were collected in the active state. When using the existing physiological characteristic signal processing methods to process these PPG signals, there are 698 PPG signals that cannot pass the signal quality verification, and 37 PPG signals that pass the signal quality verification, and the 37 sections (37/735= 80%) PPG signal is analyzed, and no false alarm occurs in the tremor warning module.

When using the physiological characteristic signal processing method shown in FIG. 6 of the present application, 588 segments (588/735≈5%) of PPG signals passed the signal quality verification, and the number of PPG signals that could not pass the signal quality verification was reduced from 698 segments to 147 segments , the recall rate of the PPG signal collected in the active state has increased by more than ten times (the signal quality pass rate has increased from 5% to 80%), and the 588-segment PPG signal is analyzed, and the atrial fibrillation warning module still has no false alarm.

Referring to FIG. 7 , it is a schematic diagram of a hardware structure of an electronic device 100 according to an embodiment of the present application. As shown in FIG. 7 , the electronic device 100 may include a processor 1001 , a memory 1002 , and a communication bus 1003 . Memory 1002 is used to store one or more computer programs 1004 . One or more computer programs 1004 are configured to be executed by the processor 1001 . The one or more computer programs 1004 include instructions that can be used to implement the above-described physiological characteristic signal processing method in the electronic device 100 .

It can be understood that the structure illustrated in this embodiment does not constitute a specific limitation on the electronic device 100 . In other embodiments, the electronic device 100 may include more or fewer components than shown, or some components may be combined, or some components may be split, or a different arrangement of components.

The processor 1001 may include one or more processing units, for example, the processor 1001 may include an application processor (application processor, AP), a graphics processor (graphics processing unit, GPU), an image signal processor (image signal processor, ISP) ), controller, video codec, DSP, CPU, baseband processor, and/or neural-network processing unit (NPU), etc. Wherein, different processing units may be independent devices, or may be integrated in one or more processors.

The processor 1001 may also be provided with a memory for storing instructions and data. In some embodiments, the memory in processor 1001 is cache memory. This memory may hold instructions or data that have just been used or recycled by the processor 1001 . If the processor 1001 needs to use the instruction or data again, it can be called directly from this memory. Repeated access is avoided, and the waiting time of the processor 1001 is reduced, thereby improving the efficiency of the system.

In some embodiments, the processor 1001 may include one or more interfaces. The interface may include an integrated circuit (inter-integrated circuit, I2C) interface, an integrated circuit built-in audio (inter-integrated circuit sound, I2S) interface, a pulse code modulation (pulse code modulation, PCM) interface, a universal asynchronous transceiver (universal asynchronous transmitter) receiver/transmitter, UART) interface, mobile industry processor interface (MIPI), general-purpose input/output (GPIO) interface, SIM interface, and/or USB interface, etc.

In some embodiments, memory 1002 may include high-speed random access memory, and may also include non-volatile memory such as hard disk, internal memory, plug-in hard disk, Smart Media Card (SMC), Secure Digital (Secure) Digital, SD) card, flash card (Flash Card), at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage device.

This embodiment also provides a computer storage medium, where computer instructions are stored in the computer storage medium, and when the computer instructions are executed on the electronic device, the electronic device executes the above-mentioned related method steps to realize the physiological characteristic signal processing in the above-mentioned embodiment. method.

This embodiment also provides a computer program product, when the computer program product runs on the computer, the computer executes the above-mentioned relevant steps, so as to realize the physiological characteristic signal processing method in the above-mentioned embodiment.

In addition, the embodiments of the present application also provide an apparatus, which may specifically be a chip, a component or a module, and the apparatus may include a connected processor and a memory; wherein, the memory is used for storing computer execution instructions, and when the apparatus is running, The processor can execute the computer-executed instructions stored in the memory, so that the chip executes the physiological characteristic signal processing methods in the foregoing method embodiments.

Wherein, the first electronic device, computer storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding method provided above. Therefore, for the beneficial effects that can be achieved, reference may be made to the provided above. The beneficial effects in the corresponding method will not be repeated here.

From the description of the above embodiments, those skilled in the art can clearly understand that for the convenience and brevity of the description, only the division of the above functional modules is used as an example for illustration. In practical applications, the above functions can be allocated as required. It is completed by different functional modules, that is, the internal structure of the device is divided into different functional modules, so as to complete all or part of the functions described above.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the device embodiments described above are only illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined. Or it may be integrated into another device, or some features may be omitted, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and components shown as units may be one physical unit or multiple physical units, that is, may be located in one place, or may be distributed to multiple different places. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of software products in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, which are stored in a storage medium , including several instructions to make a device (may be a single chip microcomputer, a chip, etc.) or a processor (processor) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program codes .

The above are only specific embodiments of the present application, but the protection scope of the present application is not limited to this, and any changes or substitutions within the technical scope disclosed in the present application should be covered within the protection scope of the present application. .

Claims

A method for processing physiological characteristic signals, comprising:

Filtering the collected physiological characteristic signal of the target individual to obtain the physiological characteristic optimization signal;

Extract peak points in the physiological characteristic optimization signal by using a preset peak extraction algorithm, and construct a first peak point set based on the extracted peak points;

The first peak point set is optimized based on the physiological characteristic signal to obtain a second peak point set, wherein the optimization process includes one of adding a peak point, deleting a peak point, and updating a peak point. or more processing; and

The physiological characteristics of the target individual are obtained by analyzing the optimized signal based on the second peak point set and the physiological characteristics.
The physiological characteristic signal processing method according to claim 1, wherein the peak point in the physiological characteristic optimization signal is extracted by using a preset peak extraction algorithm, and a first peak point set is constructed based on the extracted peak points ,include:

Use the preset peak extraction algorithm to extract the peak points in the physiological characteristic optimization signal, and construct the first peak point set based on the extracted peak points; or

The preset peak extraction algorithm is used to extract the trough points in the physiological characteristic optimization signal, and the first peak point set is constructed based on the extracted trough points.
The physiological characteristic signal processing method according to claim 1 or 2, wherein the performing optimization processing on the first peak point set based on the physiological characteristic signal comprises:

Modeling the physiological characteristic signal to obtain a physiological characteristic waveform corresponding to the physiological characteristic signal;

Splitting the physiological characteristic waveform into an up-slope waveform segment and a down-slope waveform segment, and selecting the up-slope waveform segment or the down-slope waveform segment as the target waveform;

marking each peak point in the first peak point set on the target waveform;

Divide the target waveform into a plurality of waveform windows, and calculate the initial interference degree of each of the waveform windows according to a preset interference degree calculation algorithm;

Perform optimization processing on the peak point of the waveform window, and use the preset interference degree calculation algorithm to recalculate the interference degree of the waveform window after the optimization process, until the interference degree of the waveform window achieves the minimum value, and complete an optimization process for the waveform window; and

The peak points included in each of the waveform windows that have completed the optimization process are aggregated to obtain the second peak point set.
The physiological characteristic signal processing method according to claim 3, wherein after selecting the up-slope waveform or the down-slope waveform as the target waveform, the method further comprises:

The curve segment in the target waveform is simplified to a straight line segment including only head and tail endpoints.
The physiological characteristic signal processing method according to claim 4, wherein the preset interference degree calculation algorithm comprises:

Calculate the slope distance between any two straight line segments marked with the peak point in the waveform window, and normalize the calculated slope distance;

Calculate the length ratio between any two straight line segments marked with the peak point in the waveform window, and normalize the calculated length ratio;

Calculate the absolute value of the lateral distance difference between any two straight line segments marked with the peak point in the waveform window, and normalize the calculated absolute value of the lateral distance difference;

calculating the absolute value of the longitudinal distance difference between any two straight line segments marked with the peak point in the waveform window, and normalizing the calculated absolute value of the longitudinal distance difference; and

Based on the normalized result of slope distance, the normalized result of length ratio, the normalized result of absolute value of horizontal distance difference, and the normalized result of absolute value of vertical distance difference, the interference degree of the waveform window is obtained.
The physiological characteristic signal processing method according to claim 5, wherein the normalizing the calculated slope distance comprises:

Normalizing the calculated slope distances separately, and summarizing the normalization results for each of the slope distances; or

A total slope distance is obtained by accumulating the calculated multiple slope distances, and the total slope distance is normalized.
The physiological characteristic signal processing method according to claim 5, wherein the longer straight line segment in the two straight line segments is the denominator of the length ratio, and the calculated length ratio is normalized ,include:

Normalizing the calculated length ratios separately, and summarizing the normalization results for each of the length ratios; or

The calculated length ratios are accumulated to obtain a total length ratio, and the total length ratio is normalized.
The physiological characteristic signal processing method according to claim 5, wherein the normalizing the calculated absolute value of the lateral distance difference comprises:

averaging the calculated absolute values of the plurality of lateral distance differences to obtain an average lateral distance difference; and

The calculated absolute value of each of the lateral distance differences is separately normalized based on the average lateral distance difference, and the normalized result of the absolute value of each of the lateral distance differences is summarized.
The physiological characteristic signal processing method according to claim 5, wherein the normalizing the calculated absolute value of the longitudinal distance difference comprises:

averaging the calculated absolute values of the plurality of longitudinal distance differences to obtain an average longitudinal distance difference; and

The calculated absolute value of each longitudinal distance difference is separately normalized based on the average longitudinal distance difference, and the normalized result of the absolute value of each longitudinal distance difference is summarized.
The physiological characteristic signal processing method according to claim 5, wherein the performing optimization processing on the peak point of the waveform window comprises:

Find an area where the slope distance changes abnormally, the length ratio changes abnormally, the lateral distance difference changes abnormally, or the vertical distance difference abnormally changes abnormally in the waveform window, and optimizes the peak points in the area.
The method for processing physiological characteristic signals according to claim 3, wherein the performing optimization processing on the peak point of the waveform window comprises:

When the initial interference degree of the waveform window is greater than or equal to the preset interference degree, optimization processing is performed on the peak point of the waveform window.
The physiological characteristic signal processing method according to claim 11, wherein the method further comprises:

When the initial interference degree of the waveform window is less than the preset interference degree, the optimization process for the peak point of the waveform window is abandoned.
The physiological characteristic signal processing method according to any one of claims 1 to 12, wherein the physiological characteristic of the target individual is obtained by optimizing the signal analysis based on the second peak point set and the physiological characteristic, include:

performing a signal quality assessment on the second set of peak points and the physiological characteristic optimized signal; and

When the signal quality evaluation result is that the signal quality is good, the physiological characteristics of the target individual are obtained by optimizing the signal analysis based on the second peak point set and the physiological characteristics.
A computer-readable storage medium, characterized in that, the computer-readable storage medium stores computer instructions, which, when the computer instructions are executed on an electronic device, cause the electronic device to perform the operations as in claims 1 to 13 The physiological characteristic signal processing method of any one.
An electronic device, characterized in that the electronic device comprises a processor and a memory, the memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the electronic device executes claims 1 to 1 The physiological characteristic signal processing method according to any one of claims 13.
A chip, which is coupled to a memory in an electronic device, is characterized in that, the chip is used to control the electronic device to execute the physiological characteristic signal processing method according to any one of claims 1 to 13 .