Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
In the description of the present application, it should be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or an implicit indication of the number of technical features being indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more of the described features. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
"A and/or B" includes the following three combinations: only a, only B, and combinations of a and B.
In addition, the term "plurality" in the embodiments of the present application means two or more, and in view of this, the term "plurality" may be understood as "at least two" in the embodiments of the present application. "at least one" may be understood as one or more, for example as one, two or more. For example, including at least one means including one, two or more, and not limiting what is included, e.g., including at least one of A, B and C, then A, B, C, A and B, A and C, B and C, or A and B and C, may be included.
The use of "adapted" or "configured to" in this application is meant to be open and inclusive language that does not exclude devices adapted or configured to perform additional tasks or steps. In addition, the use of "based on" is intended to be open and inclusive in that a process, step, calculation, or other action "based on" one or more of the stated conditions or values may be based on additional conditions or beyond the stated values in practice.
In this application, the term "exemplary" is used to mean "serving as an example, instance, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. The following description is presented to enable any person skilled in the art to make and use the application. In the following description, details are set forth for purposes of explanation. It will be apparent to one of ordinary skill in the art that the present application may be practiced without these specific details. In other instances, well-known structures and processes have not been shown in detail to avoid obscuring the description of the present application with unnecessary detail. Thus, the present application is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.
In a related embodiment, the method for evaluating pulse wave signal quality mainly includes:
(1) And performing pulse wave signal quality real-time evaluation by using template matching. Firstly, the pulse wave signals are required to be divided in period, secondly, a plurality of templates are established based on one period, and finally, the similarity method is used for measurement, and the signal quality is considered to be good for meeting the threshold condition. The method has the defect that a large number of templates are required to be established, and the memory consumption is high. And the detection accuracy is lowered for the inaccurate period division.
(2) And carrying out pulse wave signal quality assessment based on the frequency domain statistical characteristics. Firstly, carrying out segmentation processing on pulse wave signals, and secondly, calculating normalized autocorrelation signals of the segmented pulse wave signals; and finally, performing spectrum transformation on the autocorrelation signals, and calculating statistical characteristics of the autocorrelation signals to perform signal quality evaluation. The method can lead to greater computational complexity when evaluating in the frequency domain.
In view of the above, the present application provides a method for evaluating the quality of pulse wave signals to solve the above-mentioned problems. Referring to fig. 1, fig. 1 is a flowchart illustrating an embodiment of a pulse wave signal quality evaluation method provided in the present application, where the method includes:
step 11: a noise characteristic of the pulse wave signal to be evaluated is determined.
It will be appreciated that the signal to be evaluated in this embodiment is the original pulse wave signal acquired from the fingertip or wrist, and when the waveform is actually acquired, there are many devices and external factors to influence, which cause the waveform to change, thus causing inaccuracy of the measurement result. The interference mainly comprises: (1) instability of the light source or influence of an external light source; (2) disturbances from breathing or movement of the subject; (3) thermal noise and electromagnetic interference of the electronic device itself. The above-described disturbance causes instability or deviation of the signal at the time of signal acquisition, forming noise in the pulse wave signal.
Wherein the noise characteristic of the pulse wave signal is used to represent the case of noise contained in the pulse wave signal, alternatively the noise characteristic may be divided based on the frequency of the noise, for example the noise characteristic may comprise one or both of a high frequency noise characteristic and a low frequency noise characteristic.
Step 12: and when the noise characteristics meet the preset conditions, filtering the pulse wave signals.
Optionally, a corresponding preset condition may be preset for the noise feature, and when the noise feature meets the preset noise condition, the pulse wave signal is considered to include a noise signal or the noise signal included in the pulse wave signal may affect the pulse wave signal for subsequent blood oxygen detection.
Alternatively, in an embodiment, if the noise feature includes a high-frequency noise feature and a low-frequency noise feature, corresponding preset conditions may be set for the high-frequency noise feature and the low-frequency noise feature, respectively, for example, if the high-frequency noise feature satisfies a preset first condition, the pulse wave signal is considered to be not in accordance with a subsequent blood oxygen detection condition, or if the low-frequency noise feature satisfies a preset second condition, the pulse wave signal is considered to be not in accordance with a subsequent blood oxygen detection condition, so that filtering processing is required for the pulse wave signal in this step.
Alternatively, high frequency components may be filtered out for high frequency noise and low frequency components may be filtered out for low frequency noise.
Step 13: and determining waveform distortion characteristics of the pulse wave signals after the filtering processing.
The waveform distortion means that the signals collected in the static state are distorted in waveform and do not meet the actual requirements; if the waveform is distorted, the calculation of blood oxygen and static heart rate based on pulse wave signals is seriously affected.
Optionally, the waveform distortion characteristics may include at least one of kurtosis, skewness, and zero crossing ratio. Specifically, in practical application, one of them may be evaluated correspondingly, or two or three of them may be evaluated.
Step 14: the quality of the PPG signal is determined from the waveform distortion characteristics.
Optionally, when the kurtosis is within a preset kurtosis value range, and/or the skewness is within a preset skewness value range, and/or the zero crossing ratio is within a preset zero crossing ratio value range, determining that the pulse wave signal meets the quality requirement.
The pulse wave signal quality evaluation method provided by the embodiment comprises the following steps: determining noise characteristics of pulse wave signals to be evaluated; when the noise characteristics meet the preset conditions, filtering the pulse wave signals; determining waveform distortion characteristics of the pulse wave signals after the filtering processing; the quality of the pulse wave signal is determined from the waveform distortion characteristics. By means of the method, the quality evaluation of pulse wave signals can be achieved simply and conveniently, a large number of templates are not required to be established, the occupied memory is small, the difficulty of quality evaluation is reduced, meanwhile, the accuracy and efficiency of evaluation are further improved because periodic division and a large number of complex operations are not involved.
Referring to fig. 2, fig. 2 is a flowchart of another embodiment of a pulse wave signal quality evaluation method provided in the present application, where the method includes:
step 21: a high frequency noise characteristic of the PPG signal to be evaluated is determined, and/or a baseline drift characteristic of the PPG signal to be evaluated is determined.
The high-frequency noise interference refers to that the pulse wave signal contains more obvious high-frequency noise, and the high-frequency noise is not only one specific frequency high-frequency noise, but also may be high-frequency noise concentrated in a certain frequency interval. Reflecting that a plurality of obvious peaks appear in the frequency spectrum for the unknown noise frequency band in the frequency spectrum; in the time domain, the noise will severely affect the actual waveform, likely covering the real signal.
The low-frequency noise interference refers to the interference of a baseline drift signal, and the pulse wave signal is easily interfered by motion, respiration and equipment in the acquisition process, so that baseline drift noise is generated, that is, in the embodiment, the low-frequency noise characteristic may be a baseline drift characteristic. Too much baseline drift in the pulse wave signal can seriously affect the calculation of blood oxygen saturation, and therefore, pulse wave signals that are required to be continuously acquired (e.g., for 3 minutes) in a stationary state cannot have too much baseline drift.
Step 22: and when the high-frequency noise characteristic meets a preset first condition, filtering the PPG signal, and/or when the baseline drift characteristic meets a preset second condition, filtering the PPG signal.
It will be appreciated that the pulse wave signal cannot have more high frequency noise interference and larger baseline wander interference as a result of the analysis in step 21, and therefore the quality of the pulse wave signal can be evaluated based on at least one of the high frequency noise interference and the baseline wander interference. As an embodiment, the preset condition in step 22 may be that the characteristic value of the high frequency noise or the characteristic value of the baseline wander signal is smaller than the corresponding preset amount.
Alternatively, in an embodiment, the above-described evaluation of the high frequency noise interference and the baseline wander interference may be performed only by one of them. For example, in an application scenario, only the high-frequency noise signal is subjected to characteristic judgment, and the corresponding high-frequency noise signal in the pulse wave signal is subjected to filtering processing; in another application scenario, only the baseline wander signal is subjected to feature judgment, and the corresponding baseline wander signal in the pulse wave signal is subjected to filtering processing.
Alternatively, in another embodiment, the above-mentioned evaluation of the high-frequency noise interference and the baseline drift interference needs to be performed, and the execution sequence is not limited.
For example, in an application scenario, determining a high-frequency noise characteristic of the pulse wave signal to be evaluated, determining a baseline drift characteristic of the pulse wave signal to be evaluated when the high-frequency noise characteristic meets a preset first condition, and performing filtering processing on the pulse wave signal when the baseline drift characteristic meets a preset second condition.
Or in another application scene, determining the baseline drift characteristic of the pulse wave signal to be evaluated, determining the high-frequency noise characteristic of the pulse wave signal to be evaluated when the baseline drift characteristic meets a preset second condition, and performing filtering processing on the pulse wave signal when the high-frequency noise characteristic meets a preset first condition.
The following describes the evaluation of pulse wave signal quality from two aspects of high frequency noise and baseline wander signal, respectively.
For high frequency noise: determining a signal-to-noise ratio (SNR, signal to Noise Rario) of the pulse wave signal to be evaluated; when the signal to noise ratio is within the preset signal to noise ratio value range, filtering processing is carried out on the pulse wave signal, and when the signal to noise ratio is not within the preset signal to noise ratio value range, the pulse wave signal is considered to not meet the requirement. As one implementation mode, the data corresponding to the pulse wave signal collected in the static state can be used for calculating the signal to noise ratio at intervals of a set time period (such as N seconds), and then judging the high-frequency interference condition of the pulse wave signal according to whether the calculated signal to noise ratio is in a preset signal to noise ratio value range or not, so as to evaluate the quality of the pulse wave signal. The preset signal-to-noise ratio value range can be Th0 less than or equal to SNR less than or equal to Th1, wherein the value of Th0 is smaller than the value of Th1, and the two values can be preset or can be adjusted according to actual conditions.
Illustratively, through a number of statistical verifications, the pulse wave signal may be considered to pass the evaluation of high frequency noise interference when the signal-to-noise ratio satisfies the following inequality:
0.61≤SNR≤0.645
it can be appreciated that the preset snr range is [0.61,0.645], and can be adjusted according to a specific application scenario or a large number of experiments.
The signal-to-noise ratio calculation method is various, and the signal-to-noise ratio can be calculated only by the method, and the calculation mode is not limited in the application.
For example, the signal-to-noise ratio may be calculated using the ratio of the signal variance to the noise variance;
here, σsignal For the standard deviation, sigma, of the absolute value of the filtered pulse wave signalnoise Is the standard deviation of the noise signal after the pulse wave signal is filtered.
In this embodiment, determining the signal-to-noise ratio of the pulse wave signal to be evaluated may further include: if the signal-to-noise ratio is not in the preset signal-to-noise ratio value range, determining that the input pulse wave signal contains a high-frequency interference signal, wherein the pulse wave signal does not meet the quality requirement, and performing subsequent quality evaluation;
for baseline drift signals: determining a baseline drift amount of the pulse wave signal to be evaluated; and when the baseline drift amount is smaller than or equal to a preset drift amount threshold value, filtering the pulse wave signal. Wherein, the baseline drift amount refers to the magnitude of baseline drift. The preset drift amount threshold may be expressed as Th2, which may be preset or may be adjusted according to actual conditions.
As an implementation manner, the data corresponding to the pulse wave signal collected in the static state can be calculated at intervals of a preset time (for example, M seconds), and then the signal drift condition is judged according to whether the calculated drift is smaller than a preset drift threshold value, so as to evaluate the quality of the pulse wave signal.
In this embodiment, determining the baseline wander amount of the pulse wave signal to be evaluated may further include: if the baseline drift is greater than the preset drift threshold, the input pulse wave signal may be considered to include a baseline drift signal, which is unsatisfactory, and no further quality assessment may be performed.
Optionally, as shown in fig. 3, fig. 3 is a flowchart illustrating an embodiment of determining the baseline wander in step 22 in fig. 2, and step 22 may include:
step 221: a difference between the maximum value and the minimum value of the pulse wave signal to be evaluated is determined.
The maximum value and the minimum value of the pulse wave signal refer to the maximum value and the minimum value of the whole acquired signal. It will be appreciated that the pulse wave signal to be evaluated is typically PPG data acquired in a stationary state, typically PPG data for a preset period of time, e.g. 30s, and thus the maximum and minimum values may be those within the 30s duration.
Step 222: the median of a plurality of pulsating components corresponding to a plurality of signal segments of the pulse wave signal to be evaluated is determined.
Optionally, as shown in fig. 4, fig. 4 is a flowchart illustrating an embodiment of step 222 in fig. 3, where step 2222 may specifically include:
step 2221: a sliding time window is determined.
Step 2222: and utilizing the sliding time window and based on a set step length to slide along the time direction of the pulse wave signal so as to divide the pulse wave signal into a plurality of signal segments.
Alternatively, the time width of the sliding time window may be 2s, the moving step length of the sliding time window is 1s, for example, the divided signal segments are respectively 0-2s,1-3s,2-4s … …, that is, the pulse wave signal of 30s may be divided into 29 signal segments; in another embodiment, the time width of the sliding time window may be 2s, the moving step of the sliding time window is 2s, for example, the divided signal segments are respectively 0-2s,2-4s,4-6s … …, i.e. the pulse wave signal of 30s may be divided into 15 signal segments.
Step 2223: and determining a plurality of pulsation components corresponding to the plurality of signal terminals respectively.
As can be appreciated, the principle of pulse wave signal acquisition is: the light source (in one embodiment, including a red light source, a green light source, and an infrared light source) irradiates the skin and then reflects the light signal, which is received by the PD sensor and converted into an electrical signal to obtain a pulse wave signal. The light received by the PD sensor contains two components: one is a pulsating component (i.e., alternating current signal AC), which is an alternating component caused by light absorption of pulsating arterial blood, and the other is a stabilizing component (i.e., direct current signal DC), which reflects the magnitude of the light absorption caused by each non-pulsating tissue (e.g., epidermis, muscle, bone, vein, etc.). It is only the ratio of the amplitudes of the two-wavelength ac signal that reflects the change in blood oxygen saturation, whereas the two-wavelength dc signal can be used to scale the ac signal. Since Hb02 (oxyhemoglobin) and Hb (deoxyhemoglobin) concentrations in blood change periodically with the pulsation of blood, their absorption of light also changes pulsationally, thereby causing the intensity of the electrical signal output by the PD sensor to also change periodically with the pulsation of blood.
Step 2224: the median of the plurality of ripple components is determined.
The method for determining the median comprises the following steps: the plurality of pulsating components are ordered according to the size, if the number of the plurality of pulsating components is odd, the value of one pulsating component ordered in the middle is taken as the median, and if the number of the plurality of pulsating components is even, the average value of the values of the two pulsating components ordered in the middle is taken as the median.
The above-mentioned determination of the difference d between the maximum value and the minimum value of the pulse wave signal and the median AC of the plurality of pulse components may be specifically referred to fig. 5, and fig. 5 is a schematic diagram of the difference d between the maximum value and the minimum value of the pulse wave signal and the median AC of the plurality of pulse components.
It should be understood that the present embodiment does not limit the execution sequence of step 221 and step 222.
Step 223: and determining the baseline drift amount of the pulse wave signal to be evaluated according to the ratio of the difference value and the median.
The specific calculation formula of the baseline drift amount is as follows:
where BWD is the baseline shift amount (Baseline Wander Distance), d is the difference between the maximum and minimum values of the pulse wave signal, and AC is the median of the multiple pulsating components.
Illustratively, through a number of statistical verifications, the signal is an acceptable signal when the amount of drift satisfies the following inequality:
BWD≤4
it can be appreciated that the preset drift amount threshold value is 4, and may be adjusted according to a specific application scenario or a large number of experiments.
The step 22 further includes filtering the pulse wave signal, where the filtering may be determined according to the frequency of specific noise, and if high-frequency noise is determined in the above step, low-pass filtering is used, if low-frequency noise (baseline shift signal) is determined in the above step, high-pass filtering is used, and if both high-frequency noise and low-frequency noise are present, high-pass filtering and low-pass filtering may be used.
Step 23: and determining waveform distortion characteristics of the pulse wave signals after the filtering processing.
Waveform distortion means that signals collected in a static state are distorted in waveform and do not meet the actual requirements; if the waveform is distorted, the calculation of blood oxygen and static heart rate based on pulse wave signals is seriously affected.
Optionally, the waveform distortion characteristics may include at least one of kurtosis, skewness, and zero crossing ratio. Specifically, in practical application, one of them may be evaluated correspondingly, or two or three of them may be evaluated.
Wherein kurtosis is a statistical measure used to measure the steepness of the probability distribution of the random variable; kurtosis was determined using the following formula:
wherein, the skewness is a measure of the symmetry of the probability distribution, and the skewness is found to be related to the damaged pulse wave signals through research; the skewness is determined using the following formula:
wherein zero is the rate of change of sign describing the signal, i.e. the rate of signal from positive to negative or vice versa; the zero crossing ratio is determined using the following formula:
wherein N is the number of sampling points, xi For the signal value of the i-th sample point,is xi Mean value of σ is xi Is an indication function, { } is 1 if parameter a is true, { } is 0 if parameter a is false.
Step 24: the quality of the pulse wave signal is determined from the waveform distortion characteristics.
Optionally, when the kurtosis is within a preset kurtosis value range, and/or the skewness is within a preset skewness value range, and/or the zero crossing ratio is within a preset zero crossing ratio value range, determining that the pulse wave signal meets the quality requirement.
In a specific embodiment, the preset kurtosis value range is [1.8,2.75], the preset skewness value range is [0.2,0.85], and the preset zero crossing ratio value range is [0,0.09], namely: the three indices satisfy the following inequality simultaneously:
1.8≤K≤2.75
0.2≤S≤0.85
0≤Z≤0.09
it can be appreciated that the preset kurtosis value range, the preset skewness value range, and the preset zero-crossing ratio value range may be adjusted according to specific application scenarios or a large number of experiments, in addition to the above values.
Referring to fig. 6, fig. 6 is a flowchart of another embodiment of a pulse wave signal quality evaluation method provided in the present application, where the method includes:
(1) Inputting pulse wave signals;
(2) Calculating the signal-to-noise ratio (SNR) of the pulse wave signal;
(3) Judging whether the signal-to-noise ratio SNR meets Th 0-Th 1, and if so, performing step (4); if the pulse wave signal does not meet the requirement, determining that the input pulse wave signal contains a high-frequency interference signal, and determining that the pulse wave signal does not meet the requirement;
alternatively, the range herein may be [0.61,0.645];
(4) Calculating a baseline wander BWD of the pulse wave signal;
(5) Judging whether the baseline wander BWD meets BWD less than or equal to Th2, and if so, performing step (6); if the pulse wave signal does not meet the requirement, the input pulse wave signal comprises a baseline drift signal, and the pulse wave signal is confirmed to be inconsistent with the requirement;
alternatively, th2=4 here;
(6) Low-pass filtering and high-pass filtering are carried out on the pulse wave signals;
(7) Calculating kurtosis K, skewness S and zero crossing ratio Z of pulse wave signals;
(8) Judging whether kurtosis K, skewness S and zero crossing ratio Z meet the following conditions:
1.8≤K≤2.75
0.2≤S≤0.85
0≤Z≤0.09
if the pulse wave signal meets the requirement of signal quality, the pulse wave signal can be used for subsequent blood oxygen detection, heart rate detection, blood pressure detection and the like.
The pulse wave signal quality evaluation method provided by the embodiment comprises the following steps: determining noise characteristics of pulse wave signals to be evaluated; when the noise characteristics meet the preset conditions, filtering the pulse wave signals; determining waveform distortion characteristics of the pulse wave signals after the filtering processing; the quality of the pulse wave signal is determined from the waveform distortion characteristics. By means of the method, the quality evaluation of pulse wave signals can be achieved simply and conveniently, a large number of templates are not required to be established, the occupied memory is small, the difficulty of quality evaluation is reduced, meanwhile, the accuracy and efficiency of evaluation are further improved because periodic division and a large number of complex operations are not involved.
Referring to fig. 7, fig. 7 is a schematic structural diagram of an embodiment of a pulse wave signal quality evaluation device provided in the present application, where the pulse wave signal quality evaluation device 700 includes a processor 71 and a memory 72, where the memory 72 stores program data, and the processor 71 is configured to execute the program data to implement steps in the pulse wave signal quality evaluation method according to any one of the above-mentioned pulse wave signal quality evaluation method embodiments.
The pulse wave signal quality evaluation device 700 may be a medical device in a medical, household or wearable form, or may be an intelligent wearable device such as a wristwatch or a bracelet, wherein the processor 71 may be a processing chip, and the memory 72 may be a memory chip.
Referring to fig. 8, fig. 8 is a schematic structural diagram of an embodiment of a computer readable storage medium 800 provided in the present application, where program data 81 is stored in the computer readable storage medium, and the program data is used to implement steps in the pulse wave signal quality evaluation method according to any one of the above pulse wave signal quality evaluation method embodiments when the program data is executed by a processor.
The computer-readable storage medium 800 may be embodied in the form of a processing chip or a memory chip in the above-described embodiments.
The foregoing has outlined rather broadly the more detailed description of embodiments of the present application, wherein specific examples are provided herein to illustrate the principles and embodiments of the present application, the above examples being provided solely to assist in the understanding of the methods of the present application and the core ideas thereof; meanwhile, those skilled in the art will have variations in the specific embodiments and application scope in light of the ideas of the present application, and the present description should not be construed as limiting the present application in view of the above.