CN104887215B - Signal processing method in heart rate measurement - Google Patents
Signal processing method in heart rate measurement Download PDFInfo
- Publication number
- CN104887215B CN104887215B CN201510323890.4A CN201510323890A CN104887215B CN 104887215 B CN104887215 B CN 104887215B CN 201510323890 A CN201510323890 A CN 201510323890A CN 104887215 B CN104887215 B CN 104887215B
- Authority
- CN
- China
- Prior art keywords
- data
- heart rate
- peak
- frequency
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/024—Detecting, measuring or recording pulse rate or heart rate
- A61B5/0245—Detecting, measuring or recording pulse rate or heart rate by using sensing means generating electric signals, i.e. ECG signals
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
Abstract
The invention provides a signal processing method in heart rate measurement, which comprises the following steps: judging the acceptance or rejection of the time domain data by using a statistical method; frequency domain data, frequencies below the minimum possible heart rate being discarded, peaks in the neighborhood of the minimum possible heart rate of each main peak being discarded; the adopted frequency spectrum peaks are judged and determined one by one according to several possible frequency multiplication relations; in the case that the frequency multiplication component in the original data is small, differentiating the original data to highlight frequency multiplication; and finally, determining whether the heart rate result is adopted or not according to the quality of straight line fitting and the straight line intercept. The invention can obtain a relatively accurate heart rate value and has wide and profound application prospect in the field of heart rate measurement.
Description
Technical field
The present invention relates to medical treatment and field of signal processing, the signal processing method in more particularly to a kind of heart rate measurement.
Background technology
With the development of the society, growth in the living standard, people are increasingly paid close attention to the health status of itself；Electronics skill
Art, computer technology, the health concerns for the developing into people point of low-power consumption computing technique provide the realization rate of technology.Tradition
Medical device can provide sufficiently accurate measurement result, still, because traditional medicine apparatus needs to use in Code in Hazardous Special Locations, make
With also not comfortable enough, while measurement can not be carried out at any time.These restrictive conditions can not allow everybody in comfortable environment, the heart loosened
Feelings, health status is understood whenever and wherever possible.
Due to the development of chip technology, many low-power consumption, the processor and sensor of miniaturization are occurred in that.By at these
The collaboration of device and sensor is managed, the physiological data of many human bodies can in real time, continuously, be easily measured, such as walking, transporting
During the daily routines such as dynamic, diet, can obtain can describe the data of human body physical sign, with reference to related algorithm, can obtain
The physiological parameters such as body temperature, heart rate, blood oxygen, respiratory rate, muscle tone.Whole process can be ignored completely at these in people
Carried out in the presence of reason device and sensor.Here it is the Internet of Things industry gradually risen is in measuring of human health field
Important application.
The heart rate of people is an important indicator of human body physical sign, using frequent in clinical, motion, routine health monitoring.
At present, heart rate measurement is a very important application direction in measuring of human health field.Reliable heart rate measurement device allows
Moving with the user of usual household can be current to oneself heart rate situation have certain understanding, be conducive to
The activity or motion of progress have an anticipation.The method of heart rate measurement is a lot, can be that time-domain analysis can also be frequency domain point
Analysis.Time-domain analysis is easily limited by interference signal, and time-domain signal is gone to frequency-region signal by Fourier transform, to greatest extent
Various interference signal separation, obtain accurate heart rate signal, further, it is possible to calculate heart rate by filtering and various judgements.
In view of the above, it is necessary to provide the signal processing method in a kind of heart rate measurement, to improve rate calculation
Accuracy.
The content of the invention
The shortcoming of prior art in view of the above, at the signal in a kind of heart rate measurement
Reason method, for solving the problem of heart rate measurement is not accurate enough in the prior art.
In order to achieve the above objects and other related objects, the present invention provides the signal processing method in a kind of heart rate measurement,
The signal processing method includes step：
1) the heart rate initial data of a period of time is gathered with certain sample frequency；
2) heart rate initial data is segmented, the mean square deviation and peak-to-peak value per segment data is calculated, if certain section of mean square deviation and Feng Feng
Value is more than variance threshold values and peak-to-peak value threshold value, then the segment data is judged as into irrational data segment；
3) irrational data segment is replaced with reasonable data, the selection of reasonable data causes the data segment after replacing
There is no saltus step between its last period data segment, while all data segments ensure no saltus step below for adjustment, method is obtained according to this
The data that must be rebuild；
4) data rebuild are done with FFT, modulus obtains the frequency spectrum Value Data of domain space；
5) for the frequency spectrum Value Data of domain space, the spectrum value that frequency is less than lowest frequency value all takes 0, wherein, institute
Lowest frequency value is stated for the corresponding frequency values of human body HR min；
6) spectral peak is searched for, is arranged according to the height descending at peak, the peak after arrangement is searched for backward since first peak,
Directly delete at the peak appeared in the peak or so lowest frequency value；
7) retain three peaks of highest, and correct the position at each peak；
8) for the frequency location at three peaks of highest, its optimal frequency position relationship is judged, if optimal frequency position is closed
It is that criterion is less than judgment threshold, then exits calculating, otherwise continues to calculate；
9) fitting a straight line of frequency, and the frequency values at each peak after digital simulation are done according to optimal frequency position relationship,
The goodness of fit is less than fit threshold, or the intercept absolute value of fitting a straight line is more than intercept threshold value, then exits calculating, otherwise continue
Calculate；
10) as the frequency where the frequency location relation used is fitted three obtained peaks, digital simulation data, take respectively
Average is used as final heart rate result.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 8) in, if optimal
Frequency location relation criterion is less than judgment threshold, then goes to step a) centered difference is done to heart rate initial data and obtain new original
Heart rate data, wherein, nth data subtracts (n-1)th for (n+1)th data in old initial data in new original data
Individual data, also, using new original data as subsequent step original data；Wherein, step a) is performed in step
It is rapid 1) and step 2) between；
If performing step a)~step 8) after, optimal frequency position relationship criterion is still less than judgment threshold, then moves back
Go out to calculate, otherwise continue to calculate.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 9) in, if fitting is excellent
Degree is less than fit threshold, or the intercept absolute value of fitting a straight line is more than intercept threshold value, then goes to step a) to heart rate original number
New original data are obtained according to centered difference is done, wherein, nth data is old original number in new original data
(n+1)th data subtracts (n-1)th data in, also, using new original data as subsequent step original
Data；Wherein, step a) is performed in step 1) and step 2) between；
If performing step a)~step 9) after, the goodness of fit is still less than section of fit threshold, or fitting a straight line
Still it is more than intercept threshold value away from absolute value, then exits calculating, otherwise continues to calculate.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 1) in, minimum sampling
Frequency is met：The peak frequency for needing use is three times of the corresponding frequency of people's maximum possible heart rate, is determined according to sampling thheorem
Sample frequency be at least twice of the peak frequency.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 2) in, heart rate is original
Data sectional is according to being：Considered with the minimum possible heart rate of people, at least one complete heart rate cycle is all included per segment data.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 2) in, variance threshold values
Can set a fixed value with peak-to-peak value threshold value, can also program operation in produce in real time, or two methods synthesis,
Including：
One is initially set up than larger value, the judgement calculated every time later thinks that acceptable mean square deviation multiplies respectively
With respectively becoming new variance threshold values and peak-to-peak value threshold value after original threshold value weighted average after coefficient；Or
After each data sectional, taken from each section after the mean square deviation difference multiplying factor of minimum as threshold value, meanwhile, the threshold value
No more than one numerical value, i.e. maximum threshold value, more than the max-thresholds, then all data segments are all unacceptable, by nothing
Method carries out follow-up calculate.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 2) in, variance threshold values
It is with peak-to-peak value threshold value determination method：After each data sectional, the section of Minimum Mean Square Error is selected from each section, with the square of it
The three times of difference are variance threshold values, and nine times are peak-to-peak value threshold value.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 3) in, will be unreasonable
Data segment be replaced including：
Unreasonable data segment directly is replaced with a fixed value, the straightway after replacement does not have saltus step with front and rear data segment；
Or
Fill-error data segment is repeated with rational data segment, the data segment after replacement does not have saltus step with front and rear data segment,
If all data segments are all wrong, stop calculating, wait next batch data；If all data segments rationally if data not
Become.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 5) in, the human body
HR min is that, not less than 40 beats/min, the lowest frequency value is no more than 0.67Hz.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 7) in, using parabolic
The position at each peak of method amendment of line fitting, Gaussian function shape fitting or Cauchy-Lorenz function form fit.
It is used as a kind of preferred scheme of the signal processing method in the heart rate measurement of the present invention, step 8) in, for highest
The frequency location at three peaks, judges that its position relationship includes：1:2:3、1:3:4、1:2:4, preferentially closed with the position of last computation
System judges that acquiescence judges 1 for the first time:2:3, take optimal after judgement, be used as optimal frequency position relationship.
As described above, the present invention provides the signal processing method in a kind of heart rate measurement, have the advantages that：This hair
Bright statistical method judges the choice of time domain data；Frequency domain data, the frequency less than minimum possible heart rate is given up, each main peak
Give up at peak in minimum possible heart rate neighborhood；The spectral peak of use judges decision one by one according to possible several frequency multiplication relations,
The less situation of harmonic in initial data, to initial data difference to protrude frequency multiplication；Finally fitting a straight line is good and bad and straight
Line intercept determines whether heart rate result uses.The present invention can obtain more accurate heart rate value, have in heart rate measurement field
Extensive and far-reaching application prospect.
Brief description of the drawings
Fig. 1 is shown as the step schematic flow sheet of the signal processing method in the heart rate measurement of the present invention.
Fig. 2 is shown as in the signal processing method in the heart rate measurement of the present invention, and the position relationship of each spectral peak judges to show
It is intended to, wherein, No. 1 peak of intensity at peak is shown as in figure>No. 2 peaks>No. 3 peaks>No. 4 peaks.
Component label instructions
1 No. 1 peaks
2 No. 2 peaks
3 No. 3 peaks
4 No. 4 peaks
S1~S10 steps 1)~step 10)
Embodiment
Illustrate embodiments of the present invention below by way of specific instantiation, those skilled in the art can be by this specification
Disclosed content understands other advantages and effect of the present invention easily.The present invention can also pass through specific realities different in addition
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints with application, without departing from
Various modifications or alterations are carried out under the spirit of the present invention.
Refer to shown in Fig. 1~Fig. 2.It should be noted that the diagram provided in the present embodiment is only said in a schematic way
Bright basic conception of the invention, then in schema only display with relevant component in the present invention rather than according to component during actual implement
Number, shape and size are drawn, and kenel, quantity and the ratio of each component can be a kind of random change during its actual implementation, and
Its assembly layout kenel may also be increasingly complex.
As shown in Fig. 1~Fig. 2, the present embodiment provides the signal processing method in a kind of heart rate measurement, the signal transacting
Method includes step：
As shown in figure 1, carrying out step 1 first) S1, the heart rate initial data of a period of time is gathered with certain sample frequency.
As an example, Minimum sample rate is met：It is the corresponding frequency of people's maximum possible heart rate to need the peak frequency used
Three times of rate, the sample frequency determined according to sampling thheorem is at least twice of the peak frequency.
Specifically, because the heart rate range of people is general at 40~220bpm (number of times per minute), corresponding frequency is exactly
0.67~3.67Hz, because the present invention needs to use the second harmonic of heart rate, namely 11Hz, therefore sample frequency is at least
22Hz, or it is higher.The present invention needs to carry out FFT in low power processor, therefore data volume is as far as possible small, example
Such as, in the present embodiment, using 25Hz sample frequency, sampled data 128, about 5.12s；Between the time of rate calculation
Every less than this time, the data that calculating start time pushes away forward 5.12s participate in calculating.
As shown in figure 1, then carry out step 2) S2, heart rate initial data is segmented, calculate per segment data mean square deviation and
Peak-to-peak value, if certain section of mean square deviation and peak-to-peak value are more than variance threshold values and peak-to-peak value threshold value, the segment data is judged as unreasonable
Data segment.
As an example, heart rate initial data segmentation foundation is：Considered with the minimum possible heart rate of people, per segment data all comprising extremely
A few complete heart rate cycle.
Specifically, heart rate initial data can be divided into two sections, every section includes 16~128 data, in the present embodiment,
Every section includes 64 data.It is of course also possible to which initial data is divided into 4 sections etc., every section of data amount check included can also foundation
Demand is reasonably adjusted, however it is not limited to example recited herein.
As an example, variance threshold values and peak-to-peak value threshold value can set a fixed value, can also be real in program operation
When produce, or two methods synthesis, including：
One is initially set up than larger value, the judgement calculated every time later thinks that acceptable mean square deviation multiplies respectively
With respectively becoming new variance threshold values and peak-to-peak value threshold value after original threshold value weighted average after coefficient；Or
After each data sectional, taken from each section after the mean square deviation difference multiplying factor of minimum as threshold value, meanwhile, the threshold value
No more than one numerical value, i.e. maximum threshold value, more than the max-thresholds, then all data segments are all unacceptable, it is impossible to
Follow-up calculate will be carried out.
In the present embodiment, variance threshold values and peak-to-peak value threshold value determination method are：After each data sectional, from each section
The section of Minimum Mean Square Error is selected, using the three times of its mean square deviation as variance threshold values, nine times are peak-to-peak value threshold value
As shown in figure 1, then carrying out step 3) S3, irrational data segment is replaced with reasonable data, rationally number
According to selection cause there is no saltus step between the data segment after replacing and its last period data segment, while adjustment all data below
Section ensures no saltus step, and method obtains the data rebuild according to this.
As an example, by irrational data segment be replaced including：
Unreasonable data segment directly is replaced with a fixed value, the straightway after replacement does not have saltus step with front and rear data segment；
Or
Fill-error data segment is repeated with rational data segment, the data segment after replacement does not have saltus step with front and rear data segment,
If all data segments are all wrong, stop calculating, wait next batch data；If all data segments rationally if data not
Become.
As shown in figure 1, then carrying out step 4) data rebuild are done FFT, modulus is obtained by S4
The frequency spectrum Value Data of domain space.
As shown in figure 1, then carrying out step 5) S5, for the frequency spectrum Value Data of domain space, frequency is less than low-limit frequency
The spectrum value of value all takes 0, wherein, the lowest frequency value is the corresponding frequency values of human body HR min.
In the present embodiment, the human body HR min is that, not less than 40 beats/min, the lowest frequency value is no more than
0.67Hz, in the present embodiment, the lowest frequency value used is 0.67Hz.
As shown in figure 1, then carrying out step 6) S6, searches for spectral peak, is arranged according to the height descending at peak, after arrangement
Peak, is searched for backward since first peak, and the peak appeared in the peak or so lowest frequency value is directly deleted.
Due to the effective heart rate of highest peak general proxy in certain limit, and the HR min of general human body be not less than
40 beats/min, corresponding frequency is 0.67Hz, therefore, if occurring peak again in the range of the highest peak or so 0.67Hz, then may be used
It is determined as invalid peak, its directly deletion can be greatly improved into the accuracy of heart rate measurement.
As shown in figure 1, then carrying out step 7) S7, retain three peaks of highest, and correct the position at each peak.
As an example, can be intended using Parabolic Fit, Gaussian function shape fitting or Cauchy-Lorenz function shape
The position at each peak of method amendment of conjunction.
In the present embodiment, using Parabolic Fit each peak of method amendment position, with parabolical vertex correspondence
Frequency values be used as the revised frequency location in the peak.
As shown in Figures 1 and 2, step 8 is then carried out) S8, for the frequency location at three peaks of highest, judge its optimal frequency
Rate position relationship, if optimal frequency position relationship criterion is less than judgment threshold, exits calculating, otherwise continues to calculate.
As shown in Fig. 2 specifically, for the frequency location at three peaks of highest, judging that its position relationship includes：1:2:3、1:
3:4、1:2:4 and 2:3:4, preferentially judged with the position relationship of last computation, acquiescence judges 1 for the first time:2:3, taken after judgement
Optimal, as optimal frequency position relationship, it is 1 that Fig. 2, which show optimal position relationship,:2:3, this is a kind of relatively common position
Put putting in order for relation, i.e. highest peak and be followed successively by No. 1 peak, No. 2 peaks, No. 3 peaks and No. 4 peaks, strength relationship is No. 1 peak>No. 2
Peak>No. 3 peaks>No. 4 peaks.Certainly, due to collection or signal interference equal error, it is also possible to occur such as 1:2:4、1:3:4 situations such as,
This relation can also receive.And 2:3:4 situation is then less to be occurred.
In addition, in another more excellent embodiment, if optimal frequency position relationship criterion is less than judgment threshold, turning
Centered difference is done to heart rate initial data to step a) and obtains new original data, wherein, in new original data
Nth data subtracts (n-1)th data for (n+1)th data in old initial data, also, new original data are made
For the original data of subsequent step；Wherein, step a) is performed in step 1) and step 2) between.
If re-executing step a)~step 8) after, optimal frequency position relationship criterion is still less than judgment threshold,
Calculating is then exited, otherwise continues to calculate.
As shown in figure 1, then carrying out step 9) S9, according to optimal frequency position relationship n1:n2:The straight line that n3 does frequency is intended
Close, and the frequency values at each peak after digital simulation, intercept of the goodness of fit (GOF) less than fit threshold, or fitting a straight line
Absolute value is more than intercept threshold value, exits calculating, otherwise continues to calculate.
In addition, in another more excellent embodiment, if the goodness of fit is less than the intercept of fit threshold, or fitting a straight line
Absolute value is more than intercept threshold value, then goes to step a) and the new original data of centered difference acquisition are done to heart rate initial data,
Wherein, nth data subtracts (n-1)th data for (n+1)th data in old initial data in new original data, and
And, using new original data as subsequent step original data；Wherein, step a) is performed in step 1) and step
2) between.
If re-executing step a)~step 9) after, the goodness of fit is still less than fit threshold, or fitting a straight line
Intercept absolute value still be more than intercept threshold value, then exit calculatings, otherwise continue calculating.
As shown in figure 1, finally carrying out step 10) S10, the three peaks place obtained by the frequency location relation fitting used
Frequency, respectively digital simulation data, take average as final heart rate result.
After the present embodiment is fitted using frequency where multiple peaks, then the heart rate result that the method being averaged is obtained, tool
There is higher reliability, the accuracy of heart rate measuring and calculating can be greatly improved.
As described above, the present invention provides the signal processing method in a kind of heart rate measurement, the signal processing method includes
Step：1) the heart rate initial data of a period of time is gathered with certain sample frequency；2) heart rate initial data is segmented, calculates every section
The mean square deviation and peak-to-peak value of data, if certain section of mean square deviation and peak-to-peak value are more than variance threshold values and peak-to-peak value threshold value, by the hop count
It is judged that being irrational data segment；3) irrational data segment is replaced with reasonable data, the selection of reasonable data makes
There is no saltus step between data segment and its last period data segment after must replacing, while all data segments ensure do not have below for adjustment
Saltus step, the data that method acquisition is rebuild according to this；4) data rebuild are done with FFT, modulus obtains frequency
The frequency spectrum Value Data of domain space；5) for the frequency spectrum Value Data of domain space, the spectrum value that frequency is less than lowest frequency value is whole
0 is taken, wherein, the lowest frequency value is the corresponding frequency values of human body HR min；6) spectral peak is searched for, according to the high sinking at peak
Sequence is arranged, and the peak after arrangement is searched for backward since first peak, and the peak appeared in the peak or so lowest frequency value is directly deleted
Remove；7) retain three peaks of highest, and correct the position at each peak；8) for the frequency location at three peaks of highest, judge it most
Excellent frequency location relation, if optimal frequency position relationship criterion is less than judgment threshold, goes to step a) to heart rate original number
New original data are obtained according to centered difference is done, wherein, nth data is old original number in new original data
(n+1)th data subtracts (n-1)th data in, also, using new original data as subsequent step original
Data；Wherein, step a) is performed in step 1) and step 2) between；9) straight line for doing frequency according to optimal frequency position relationship is intended
Close, and the frequency values at each peak after digital simulation, if the goodness of fit is absolute less than fit threshold, or the intercept of fitting a straight line
Value is more than intercept threshold value, then goes to step a) and the new original data of centered difference acquisition are done to heart rate initial data, wherein,
Nth data subtracts (n-1)th data for (n+1)th data in old initial data in new original data, also, will
New original data as subsequent step original data；Wherein, step a) is performed in step 1) and step 2) it
Between；10) as the frequency where the frequency location relation used is fitted three obtained peaks, digital simulation data, take average respectively
It is used as final heart rate result.The present invention judges the choice of time domain data with statistical method；Frequency domain data, less than minimum possible heart rate
Frequency give up, the peak that the minimum of each main peak may be in heart rate neighborhood is given up；The spectral peak of use is according to possible several times
Frequency relation judges to determine one by one；The less situation of harmonic in initial data, to initial data difference to protrude frequency multiplication；Most
The quality and Linear intercept of fitting a straight line determine whether heart rate result uses afterwards.The present invention can obtain more accurate heart rate
Value, has extensive and far-reaching application prospect in heart rate measurement field.So, the present invention effectively overcomes kind of the prior art
Plant shortcoming and have high industrial utilization.
The above-described embodiments merely illustrate the principles and effects of the present invention, not for the limitation present invention.It is any ripe
Know the personage of this technology all can carry out modifications and changes under the spirit and scope without prejudice to the present invention to above-described embodiment.Cause
This, those of ordinary skill in the art is complete without departing from disclosed spirit and institute under technological thought such as
Into all equivalent modifications or change, should by the present invention claim be covered.
Claims (10)
1. the signal processing method in a kind of heart rate measurement, it is characterised in that the signal processing method includes step：
1) the heart rate initial data of a period of time is gathered with certain sample frequency；
2) heart rate initial data is segmented, the mean square deviation and peak-to-peak value per segment data is calculated, if certain section of mean square deviation and peak-to-peak value are big
In variance threshold values and peak-to-peak value threshold value, then the segment data is judged as irrational data segment；
3) irrational data segment is replaced with reasonable data, the selection of reasonable data causes the data segment after replacing and it
The last period data segment between there is no saltus step, while adjustment below all data segments ensure no saltus step, method is weighed according to this
The data newly built；
4) data rebuild are done with FFT, modulus obtains the frequency spectrum Value Data of domain space；
5) for the frequency spectrum Value Data of domain space, frequency be less than lowest frequency value spectrum value all take 0, wherein, it is described most
Low frequency value be the corresponding frequency values of human body HR min, the human body HR min be not less than 40 beats/min, it is described minimum
Frequency values are no more than 0.67Hz；
6) spectral peak is searched for, arranges, the peak after arrangement is searched for backward since first peak, occurs according to the height descending at peak
Directly delete at peak in the peak or so lowest frequency value；
7) retain three peaks of highest, and correct the position at each peak；
8) for the frequency location at three peaks of highest, its optimal frequency position relationship is judged, if optimal frequency position relationship is sentenced
According to less than judgment threshold, then calculating is exited, otherwise continue to calculate；
9) fitting a straight line of frequency, and the frequency values at each peak after digital simulation are done according to optimal frequency position relationship, are fitted
Goodness is less than fit threshold, or the intercept absolute value of fitting a straight line is more than intercept threshold value, then exits calculating, otherwise continues to count
Calculate；
10) as the frequency where the frequency location relation used is fitted three obtained peaks, digital simulation data, take average respectively
It is used as final heart rate result.
2. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 8) in, if most
Excellent frequency location relation criterion is less than judgment threshold, then goes to step a) and the new original of centered difference acquisition is done to heart rate initial data
Beginning heart rate data, wherein, nth data is that (n+1)th data subtracts the in old initial data in new original data
N-1 data, also, using new original data as subsequent step original data；Wherein, step a) is performed
Step 1) and step 2) between, n is positive integer；
If performing step a)~step 8) after, optimal frequency position relationship criterion is still less than judgment threshold, then exits meter
Calculate, otherwise continue to calculate.
3. the signal processing method in heart rate measurement according to claim 1 or 2, it is characterised in that：Step 9) in, if intending
Close goodness and be less than fit threshold, or the intercept absolute value of fitting a straight line is more than intercept threshold value, then goes to step a) former to heart rate
Beginning data do centered difference and obtain new original data, wherein, nth data is old original in new original data
(n+1)th data subtracts (n-1)th data in beginning data, also, regard new original data as the original of subsequent step
Heart rate data；Wherein, step a) is performed in step 1) and step 2) between, n is positive integer；
If performing step a)~step 9) after, the goodness of fit is still exhausted less than fit threshold, or the intercept of fitting a straight line
Intercept threshold value is still more than to value, then exits calculating, otherwise continues to calculate.
4. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 1) in, it is minimum to adopt
Sample frequency is met：The peak frequency for needing use is three times of the corresponding frequency of people's maximum possible heart rate, is determined according to sampling thheorem
Fixed sample frequency is at least twice of the peak frequency.
5. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 2) in, heart rate is former
Beginning data sectional is according to being：Considered with the minimum possible heart rate of people, at least one complete heart rate cycle is all included per segment data.
6. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 2) in, variance threshold
Value and peak-to-peak value threshold value can set a fixed value, can also in real time be produced in program operation, or two methods is comprehensive
Close, including：
One is initially set up than larger value, the judgement calculated every time later thinks that acceptable mean square deviation distinguishes multiplying factor
Afterwards with respectively becoming new variance threshold values and peak-to-peak value threshold value after original threshold value weighted average；Or
After each data sectional, taken from each section after the mean square deviation difference multiplying factor of minimum as threshold value, meanwhile, the threshold value can not
More than one numerical value, i.e. maximum threshold value, more than the max-thresholds, then all data segments are all unacceptable, it is impossible to will be entered
Row is follow-up to be calculated.
7. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 2) in, variance threshold
Value and peak-to-peak value threshold value determination method are：After each data sectional, the section of Minimum Mean Square Error is selected from each section, with the equal of it
The three times of variance are variance threshold values, and nine times are peak-to-peak value threshold value.
8. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 3) in, it will not conform to
The data segment of reason be replaced including：
Unreasonable data segment directly is replaced with a fixed value, the straightway after replacement does not have saltus step with front and rear data segment；Or
Fill-error data segment is repeated with rational data segment, the data segment after replacement does not have saltus step with front and rear data segment, if
All data segments are all wrong, then stop calculating, wait next batch data；Data are constant if all data segments are reasonable.
9. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 7) in, using throwing
The position at each peak of method amendment of the fitting of thing line, Gaussian function shape fitting or Cauchy-Lorenz function form fit.
10. the signal processing method in heart rate measurement according to claim 1, it is characterised in that：Step 8) in, for most
The frequency location at high three peaks, judges that its position relationship includes：1:2:3、1:3:4、1:2:4, with the position relationship of last computation
Judge, acquiescence judges 1 for the first time:2:3, take optimal after judgement, be used as optimal frequency position relationship.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510323890.4A CN104887215B (en) | 2015-06-12 | 2015-06-12 | Signal processing method in heart rate measurement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510323890.4A CN104887215B (en) | 2015-06-12 | 2015-06-12 | Signal processing method in heart rate measurement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104887215A CN104887215A (en) | 2015-09-09 |
| CN104887215B true CN104887215B (en) | 2017-11-03 |
Family
ID=54020499
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510323890.4A Active CN104887215B (en) | 2015-06-12 | 2015-06-12 | Signal processing method in heart rate measurement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104887215B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106383808B (en) * | 2016-09-18 | 2019-08-02 | 时瑞科技（深圳）有限公司 | The processing system and method for heart rate electrocardiosignal |
| CN106872021A (en) * | 2017-02-24 | 2017-06-20 | 广东电网有限责任公司江门供电局 | A kind of method of distributed power transmission line vibrational spectra information extraction |
| CN107960992B (en) * | 2017-03-16 | 2023-12-01 | 纳智源科技（唐山）有限责任公司 | Signal processing device and method based on human respiratory heartbeat characteristic wave |
| EP3476282B1 (en) | 2017-08-22 | 2020-11-25 | Shenzhen Goodix Technology Co., Ltd. | Heart rate measuring method and apparatus, and electronic terminal |
| CN109330580A (en) * | 2018-11-22 | 2019-02-15 | 深圳市元征科技股份有限公司 | A kind of rhythm of the heart method, system and associated component |
| CN114246579B (en) * | 2020-09-23 | 2024-03-15 | 深圳绿米联创科技有限公司 | Heart rate value determining method and device, terminal equipment and storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE10006154A1 (en) * | 2000-02-11 | 2001-08-16 | Biotronik Mess & Therapieg | Process for calculating the heart rate variability for use in an EKG monitor and EKG monitor with a corresponding calculation program |
| EP2327361A1 (en) * | 2006-03-03 | 2011-06-01 | Cardiac Science Corporation | Methods for quantifying the risk of cardiac death using exercise induced heart rate variability metrics |
| US20130085354A1 (en) * | 2007-01-10 | 2013-04-04 | Starr Life Sciences Corp. | Techniques for accurately deriving physiologic parameters of a subject from photoplethysmographic measurements |
| US8740803B2 (en) * | 2010-03-23 | 2014-06-03 | General Electric Company | Use of the frequency spectrum of artifact in oscillometry |
-
2015
- 2015-06-12 CN CN201510323890.4A patent/CN104887215B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN104887215A (en) | 2015-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104887215B (en) | Signal processing method in heart rate measurement | |
| US10679757B2 (en) | Method and apparatus for establishing a blood pressure model and method and apparatus for determining a blood pressure | |
| CN106971059B (en) | Wearable equipment based on neural network self-adaptation health monitoring | |
| Tan et al. | Non-invasive continuous blood pressure measurement based on mean impact value method, BP neural network, and genetic algorithm | |
| US20160223514A1 (en) | Method for denoising and data fusion of biophysiological rate features into a single rate estimate | |
| CN105167742B (en) | A kind of fetal weight adaptive estimation method and system | |
| CN109461188B (en) | A kind of two-dimensional x-ray cephalometry image anatomical features point automatic positioning method | |
| CN113349752B (en) | Wearable device real-time heart rate monitoring method based on sensing fusion | |
| CN110309817A (en) | A kind of pulse wave motion artifacts minimizing technology of parameter adaptive optimization VMD | |
| CN113171080B (en) | Energy metabolism evaluation method and system based on wearable sensing information fusion | |
| CN106691402A (en) | Fatigue level analysis method and device based on pulse characteristics | |
| CN107411702A (en) | A kind of method and system for testing Wrist wearable type terminal heart rate detection precision | |
| CN114628033A (en) | Disease risk prediction method, device, equipment and storage medium | |
| CN104757938A (en) | Near infrared spectrum technology based skin lipid content detection method | |
| Roy et al. | BePCon: A photoplethysmography-based quality-aware continuous beat-to-beat blood pressure measurement technique using deep learning | |
| CN112768074A (en) | Artificial intelligence-based serious disease risk prediction method and system | |
| CN114098658A (en) | Health state monitoring method and device | |
| CN106580324A (en) | Method and device for extracting respiratory signal | |
| CN116701886B (en) | Intelligent analysis management system for exercise lung test data | |
| CN116701977A (en) | Human body temperature data fitting method based on clustering algorithm and neural network | |
| CN111755127A (en) | Blood glucose estimation method based on metabolic mixing method | |
| CN116451110A (en) | Blood glucose prediction model construction method based on signal energy characteristics and pulse period | |
| CN114468996B (en) | Method for analyzing breast signs based on order, multi-mode and symmetry deficiency | |
| CN107564588B (en) | Physiological health data prediction device | |
| CN114145725B (en) | PPG sampling rate estimation method based on noninvasive continuous blood pressure measurement |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |