CN117033906A - BCG signal heart rate extraction method based on multidimensional features - Google Patents
BCG signal heart rate extraction method based on multidimensional features Download PDFInfo
- Publication number
- CN117033906A CN117033906A CN202311010661.8A CN202311010661A CN117033906A CN 117033906 A CN117033906 A CN 117033906A CN 202311010661 A CN202311010661 A CN 202311010661A CN 117033906 A CN117033906 A CN 117033906A
- Authority
- CN
- China
- Prior art keywords
- heart rate
- bpm
- data
- steps
- feature
- 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.)
- Pending
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 claims abstract description 23
- 238000001914 filtration Methods 0.000 claims abstract description 13
- 238000005311 autocorrelation function Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 4
- 230000003321 amplification Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 abstract description 5
- 239000000284 extract Substances 0.000 description 6
- 230000007547 defect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 2
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 2
- 230000010247 heart contraction Effects 0.000 description 2
- 239000002033 PVDF binder Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 230000033001 locomotion Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229920002981 polyvinylidene fluoride Polymers 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- 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
-
- 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
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Heart & Thoracic Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Veterinary Medicine (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Physiology (AREA)
- Cardiology (AREA)
- Psychiatry (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Databases & Information Systems (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Evolutionary Biology (AREA)
- Computational Mathematics (AREA)
- Algebra (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Power Engineering (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a BCG signal heart rate extraction method based on multidimensional features; the method comprises the following steps: filtering the X to obtain a filtered BCG signal X f The method comprises the steps of carrying out a first treatment on the surface of the X is extracted by using a nonlinear differential enhancement method f Non-linear differential enhancement feature set X of (2) fd The method comprises the steps of carrying out a first treatment on the surface of the For X f Extracting a wide time scale feature set X fw The method comprises the steps of carrying out a first treatment on the surface of the For X f Extracting narrow time scale feature set X fs The method comprises the steps of carrying out a first treatment on the surface of the Extracting X f Feature set X of two frequency bands fl And X fh The method comprises the steps of carrying out a first treatment on the surface of the The invention has the characteristics of strong anti-interference capability to noise, low heart rate error of final calculation, high accuracy and good practicability.
Description
Technical Field
The invention relates to the technical field of heart rate extraction, in particular to a BCG signal heart rate extraction method with high accuracy, good robustness and strong anti-interference capability based on multidimensional features.
Background
In recent years, the heart rate index extraction method has important application in the fields of professional medical diagnosis, intelligent home systems, personal health care and the like. The heart rate extraction method based on BCG collects pressure changes in the human heart contraction process through the PVDF piezoelectric film sensor, so that the heart rate extraction is realized. The heart rate extraction method based on the BCG signals can collect signals through the sensor placed at the position near the chest below the bed sheet, has no wearing requirement and no battery power problem, and is suitable for long-time sleeping at night.
However, in the current use environment, the actual effect of the BCG signal is not good, and the main reason is that the current BCG method only extracts the heart rate based on a single feature, so that the algorithm is too sensitive to the single feature, and the feature parameter weight is too large. When the characteristics are greatly affected by noise or individual signal differences, the overall algorithm performance may be significantly reduced. In addition, during long-term heart rate extraction, BCG signals may be disturbed by body movement, respiration, environmental noise, and the like. Therefore, the current BCG heart rate extraction method based on the single feature method is low in accuracy, poor in robustness and insufficient in practicability.
Disclosure of Invention
The invention aims to overcome the defects of low accuracy and poor robustness of a BCG signal heart rate extraction method based on a single feature method in the prior art, and provides a BCG signal heart rate extraction method based on multidimensional features, which has high accuracy, good robustness and strong anti-interference capability.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a multi-dimensional characteristic-based BCG signal heart rate extraction method comprises the steps that a plurality of sections of BCG signals with the length of n, which are arranged according to time sequence, are stored in a memory, and a processor performs data processing on the BCG signals X of the M-th section, wherein M is more than or equal to t w /n+1,t w Is a wide time scale coefficient; the method comprises the following steps:
step 1, the filtering treatment is carried out on X,obtaining a BCG signal X after filtering f :
X={x(1),x(2),x(3),…,x(n)};
X f =filter(X)={x f (1),x f (2),x f (3),…,x f (n)};
The processor filters X by using a fourth-order Butterworth filter with the bandwidth of 0.5Hz-9Hz, filters out signal data of a non-heart rate frequency band, eliminates noise and obtains a BCG signal X after the filtering treatment f ;
Step 2, extracting X by using a nonlinear differential enhancement method f Non-linear differential enhancement feature set X of (2) fd Wherein X is fd ={x fd (1),x fd (2),x fd (3),…,x fd (n)};
Step 3, for X f Extracting a wide time scale feature set X fw :
x fw (i)=x f (i-t w );
X fw ={x fw (1),x fw (2),x fw (3),…,x fw (n)};
Wherein x is fw (i) Is X fw Elements of (a) and (b);
step 4, for X f Extracting narrow time scale feature set X fs :
x fs (i)=x f (i-t s );
X fs ={x fs (1),x fs (2),x fs (3),…,x fs (n)};
Wherein x is fs (i) Is X fs Elements of (a) and (b);
step 5, using two fourth order Butterworth low pass filters to make X f Filtering respectively, and extracting X with cut-off frequencies of 2Hz and 4Hz of two fourth-order Butterworth low-pass filters f Feature set X of two frequency bands fl And X fh ;
Wherein X is fl ={x fl (1),x fl (2),x fl (3),…,x fl (n)};
X fh ={x fh (1),x fh (2),x fh (3),…,x fh (n)};
x fl (i) Is X fl Element x of (a) fh (i)X fh Elements of (a) and (b);
step 6, calculating an autocorrelation function F corresponding to the feature space F (i) based on the multidimensional features r (i);
Step 7, find F r (i) Determining heart rates respectively corresponding to the multidimensional features by using the maximum peak value in the peak set;
and 8, placing H into different heart rate intervals, selecting one or two heart rate intervals with the largest amount of heart rate data, calculating the average value of all heart rate data in all selected heart rate intervals, and taking the average value as the obtained final heart rate.
Compared with a single-feature-based BCG heart rate extraction method, the multi-dimensional features of the BCG signals are more comprehensively analyzed and extracted in the steps 2-5; even if a certain characteristic is invalid due to noise influence or individual difference of users, the invention can still extract other effective characteristics, and the invention respectively extracts nonlinear enhancement characteristics, different time scale characteristics and different frequency band characteristics in the BCG signal. The nonlinear enhancement features effectively extract real-time nonlinear variation information of the BCG signals, the different time scale features extract trend information of the BCG signals based on previous signal paragraphs, and the different frequency band features extract real-time variation information of the BCG signals based on different frequency bands. Meanwhile, the heart rate calculation method based on the autocorrelation function in the steps 6-7 can further avoid interference of harmonic components and noise in the characteristic signals, and heart rate values can be effectively calculated and extracted.
Therefore, the method effectively reduces the influence of individual difference of the BCG signals on the heart rate calculation accuracy, enhances the anti-interference capability of the algorithm, improves the heart rate extraction accuracy, and simultaneously ensures that the performance of the algorithm is more stable.
Preferably, step 2 comprises the steps of:
step 2-1, pairX f Performing first-order median difference processing to obtain median difference characteristics x d (i):
x d (i1)=(x f (i1-1)-x f (i1+1))/2,i1=2,3,…,n-1;
x d (1)=x f (1),x d (n)=x f (n),i=1,2,3,…,n;
Step 2-2, for x d (i) Scaling is performed:
find x d (1),x d (2),x d (3),…,x d Maximum value x in (n) dmax By usingCalculating x d average ;
Using the formula alpha x d (i)/(x dmax ×x d average ) For x d (i) Scaling, wherein alpha is a scaling parameter, and alpha is more than or equal to 1 and less than or equal to 8;
step 2-3, performing feature amplification by using a nonlinear mapping function to obtain nonlinear differential enhancement features X fd :
X fd ={x fd (1),x fd (2),x fd (3),…,x fd (n)}。
Preferably, step 6 comprises the steps of:
the autocorrelation function F corresponding to the feature space F (i) based on the multidimensional feature is calculated by using the following formula r (i):
Wherein F (i) = { x fd (i),x fw (i),x fs (i),x fl (i),x fh (i)};
F r (i)={x rfd (i),x rfw (i),x rfs (i),x rfl (i),x rfh (i)};
k=0,1,2,…,n-1;
Preferably, step 7 comprises the steps of:
F r (i) Each element of the set comprises n data, and the n data of each element are subjected to normalization processing to obtain five sets Fr1, fr2, fr3, fr4 and Fr5 with the length of n: fr 1= { Fr1 (1), fr1 (2), fr1 (3), …, fr1 (n) };
Fr2={Fr2(1),Fr2(2),Fr2(3),…,Fr2(n)};
Fr3={Fr3(1),Fr3(2),Fr3(3),…,Fr3(n)};
Fr4={Fr4(1),Fr4(2),Fr4(3),…,Fr4(n)};
Fr5={Fr5(1),Fr5(2),Fr5(3),…,Fr5(n)};
the sets P1, P2, P3, P4 and P5 of all the peak points of Fr1, fr2, fr3, fr4, fr5 are determined respectively using the following formulas:
p1= { i1|fr1 (i 1) > Fr1 (i 1-1) and F r1 (i 1) > Fr1 (i 1+1) };
p2= { i1|fr2 (i 1) > F r2 (i 1-1) and F r2 (i 1) > Fr2 (i1+1) };
p3= { i1|fr3 (i 1) > F r3 (i 1-1) and F r3 (i 1) > Fr3 (i1+1) };
p4= { i1|fr4 (i 1) > F r (i 1-1) and F r4 (i 1) > Fr4 (i1+1) };
p5= { i1|fr5 (i 1) > F r5 (i 1-1) and F r5 (i 1) > Fr5 (i1+1) };
let j be the number of the set of all peak points, j=1, 2,3,4,5; since the peak set Pj of the autocorrelation function represents different time domain components of the data, the largest peak value Pj in Pj max Instantaneous heart rate H, which is a feature corresponding to Pj j The processor calculates and obtains an instantaneous heart rate set H:
H={H 1 ,H 2 ,H 3 ,H 4 ,H 5 store H into memory.
Preferably, step 8 comprises the steps of:
setting heart rate intervals which are sequentially arranged from low to high: [30, 35) bpm, [35, 40) bpm, [40, 45) bpm, [45, 50) bpm, [50, 55) bpm, …, [165, 170) bpm, [170, 175) bpm, [175,180] bpm;
the processor will H in H 1 ,H 2 ,H 3 ,H 4 ,H 5 And respectively classifying the heart rate intervals with the values into different heart rate intervals, selecting one or two heart rate intervals with the largest quantity of heart rate data, calculating the average value of all heart rate data in all selected intervals, and taking the average value as the obtained final heart rate.
Therefore, the invention has the following beneficial effects: the data processing is more accurate and effective, the calculated amount is smaller, the calculation speed is high, the anti-interference capability on noise is strong, the heart rate error of final calculation is small, the accuracy is high, the defect of an extraction method based on single characteristics is overcome, and meanwhile the practicability is improved.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic representation of X of the present invention;
FIG. 3 is an X of the present invention f Is a schematic diagram of (a);
FIG. 4 is an X of the present invention fd Is a schematic diagram of (a);
fig. 5 is a schematic diagram of heart rate data distribution according to an embodiment of the present invention.
Detailed Description
The invention is further described below with reference to the drawings and detailed description.
The embodiment shown in fig. 1 is a BCG signal heart rate extraction method based on multidimensional features, wherein a plurality of segments of BCG signals with a length of n=500, which are arranged according to time sequence, are stored in a memory, and a processor performs data processing on BCG signals X of the 8 th segment; the method comprises the following steps:
step 1, performing filtering processing on X shown in FIG. 2 to obtain a filtered BCG signal X f :
X={x(1),x(2),x(3),…,x(n)};
X f =filter(X)={x f (1),x f (2),x f (3),…,x f (n)};
The processor filters X by a fourth-order Butterworth filter with the bandwidth of 0.5Hz-9Hz, filters out signal data of non-heart rate frequency bands, eliminates noise, and obtains a BCG signal X after the filtering processing as shown in figure 3 f The method comprises the steps of carrying out a first treatment on the surface of the N=500 in the present embodiment.
Step 2, extracting X by using a nonlinear differential enhancement method f Non-linear differential enhancement feature set X of (2) fd Wherein X is fd ={x fd (1),x fd (2),x fd (3),…,x fd (n)}:
The BCG signal is essentially a pressure signal generated by heart beating, firstly, a differential function method is utilized to search dynamic change characteristics in the pressure signal, then a scaling function method is utilized to further strengthen the characteristics, and finally, a non-linear mapping function method is utilized to map the characteristics to an exponential domain to further amplify the characteristics.
Step 2-1, for X f Performing first-order median difference processing to obtain median difference characteristics x d (i):
x d (i1)=(x f (i1-1)-x f (i1+1))/2,i1=2,3,…,n-1;
x d (1)=x f (1),x d (n)=x f (n),i=1,2,3,…,n;
Step 2-2, for x d (i) Scaling is performed:
find x d (1),x d (2),x d (3),…,x d Maximum value x in (n) dmax By usingCalculating x d average ;
Using the formula alpha x d (i)/(x dmax ×x d average ) For x d (i) Scaling, wherein alpha is a scaling parameter, and alpha=4;
step 2-3, performing feature amplification by using the nonlinear mapping function to obtain nonlinear differential enhancement features X shown in FIG. 4 fd :
X fd ={x fd (1),x fd (2),x fd (3),…,x fd (n)};
Step 3, for X f At a wide time scale coefficient t w Under the condition of=2500, a wide time scale feature set X is extracted fw :
x fw (i)=x f (i-t w );
X fw ={x fw (1),x fw (2),x fw (3),…,x fw (n)};
The processor performs filtering processing on each section of BCG signals before X in advance, and the obtained filtered data results of each section of BCG signals are stored in the memory.
Extracting the data of the data segment before the current data segment by using the formula, and x f (i) Is the data of the 8 th segment BCG signal, x fw (i) Is that8-2500/500=3, data in the 3 rd segment BCG signal.
Step 4, for X f At a narrow time scale coefficient t s Under the condition of=625, a narrow time scale feature set X is extracted fs :
x fs (i)=x f (i-t s );
X fs ={x fs (1),x fs (2),x fs (3),…,x fs (n)};
Extracting the data of the data segment before the current data segment by using the formula, and x f (i) Is the data of the 8 th segment BCG signal, x fs (i) Is with x f (i) Data with a distance 625 between them.
Step 5, using two fourth order Butterworth low pass filters to make X f Filtering respectively, and extracting X with cut-off frequencies of 2Hz and 4Hz of two fourth-order Butterworth low-pass filters f Feature set X of two frequency bands fl And X fh ;X fl Corresponding to a low pass filter with a cut-off frequency of 2 Hz; x is X fh Corresponding to a low pass filter with a cut-off frequency of 4 Hz;
wherein X is fl ={x fl (1),x fl (2),x fl (3),…,x fl (n)};
X fh ={x fh (1),x fh (2),x fh (3),…,x fh (n)};
x fl (i) Is X fl Element x of (a) fh (i)X fh Elements of (a) and (b);
step 6, calculating the autocorrelation function F corresponding to the feature space F (i) based on the multidimensional feature by using the following formula r (i);
Wherein F (i) = { x fd (i),x fw (i),x fs (i),x fl (i),x fh (i)};
F r (i)={x rfd (i),x rfw (i),x rfs (i),x rfl (i),x rfh (i)};
k=0,1,2,…,n-1;
Step 7,F r (i) Each element of the set comprises n data, and the n data of each element are subjected to normalization processing to obtain five sets Fr1, fr2, fr3, fr4 and Fr5 with the length of n:
Fr1={Fr1(1),Fr1(2),Fr1(3),…,Fr1(n)};
Fr2={Fr2(1),Fr2(2),Fr2(3),…,Fr2(n)};
Fr3={Fr3(1),Fr3(2),Fr3(3),…,Fr3(n)};
Fr4={Fr4(1),Fr4(2),Fr4(3),…,Fr4(n)};
Fr5={Fr5(1),Fr5(2),Fr5(3),…,Fr5(n)};
the sets P1, P2, P3, P4 and P5 of all the peak points of Fr1, fr2, fr3, fr4, fr5 are determined respectively using the following formulas:
p1= { i1|fr1 (i 1) > Fr1 (i 1-1) and F r1 (i 1) > Fr1 (i 1+1) };
p2= { i1|fr2 (i 1) > F r2 (i 1-1) and F r2 (i 1) > Fr2 (i1+1) };
p3= { i1|fr3 (i 1) > F r3 (i 1-1) and F r3 (i 1) > Fr3 (i1+1) };
p4= { i1|fr4 (i 1) > F r (i 1-1) and F r4 (i 1) > Fr4 (i1+1) };
p5= { i1|fr5 (i 1) > F r5 (i 1-1) and F r5 (i 1) > Fr5 (i1+1) };
let j be the number of the set of all peak points, j=1, 2,3,4,5; since the peak set Pj of the autocorrelation function represents different time domain components of the data, the largest peak value Pj in Pj max Instantaneous heart rate H, which is a feature corresponding to Pj j The processor calculates and instantaneous heart rate set H: h= { H 1 ,H 2 ,H 3 ,H 4 ,H 5 Storing H into memory;
and 8, placing H into different heart rate intervals, selecting one or two heart rate intervals with the largest amount of heart rate data, calculating the average value of all heart rate data in all selected heart rate intervals, and taking the average value as the obtained final heart rate (unit bpm/heart beat per minute).
Setting heart rate intervals which are sequentially arranged from low to high: [30, 35) bpm, [35, 40) bpm, [40, 45) bpm, [45, 50) bpm, [50, 55) bpm, …, [165, 170) bpm, [170, 175) bpm, [175,180] bpm;
the processor will H of all segments of BCG data 1 ,H 2 ,H 3 ,H 4 ,H 5 And respectively classifying the heart rate intervals with the values into different heart rate intervals, selecting one or two heart rate intervals with the largest quantity of heart rate data, calculating the average value of all heart rate data in all selected intervals, and taking the average value as the obtained final heart rate.
As shown in fig. 5, the heart rate data set h= { H 1 =60,H 2 =61,H 3 =63,H 4 =78,H 5 =130 }. First, H is 1 ,H 2 ,H 3 Falls within the interval [60,65 ], H is taken as 4 Falls within the interval [75,80 ], H 5 Falls under [130,135 ] regionIn the compartment. Of the 3 intervals, the heart rate data amount of the [60, 65) interval has the highest weight, accounting for 60% of the total heart rate data amount. Thus, the final heart rate is calculated using heart rate data H in the range of 60 to 65 1 ,H 2 ,H 3 To calculate the average value, resulting in 61.3, leaving the decimal point one bit behind.
The table shown below is a prior art waveform based heart rate extraction method based on error comparison of a single feature heart rate extraction method with the present invention heart rate extraction method. Wherein, an average absolute error (MAE) and an average absolute percent error (MAPE) of the heart rate and the standard reference heart rate are output as error statistical parameters by using an algorithm.
Compared with the prior art, the heart rate extraction method provided by the invention is improved by 46.9% on average in the MAE aspect and is improved by 49.7% on average in the MAPE aspect. The data processed by the method is more accurate and effective, has stronger noise immunity, and has lower heart rate error and higher accuracy. Meanwhile, the invention overcomes the defects of the traditional BCG heart rate extraction method and remarkably improves the practicability of the BCG heart rate extraction.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.
Claims (5)
1. A BCG signal heart rate extraction method based on multidimensional features is characterized in that a plurality of sections of BCG signals with the length of n, which are arranged according to time sequence, are stored in a memory, and a processor performs data processing on BCG signals X of an M-th section, wherein M is more than or equal to t w /n+1,t w Is a wide time scale systemA number; the method comprises the following steps:
step 1, performing filtering treatment on X to obtain a BCG signal X after the filtering treatment f :
X={x(1),x(2),x(3),…,x(n)};
X f =filter(X)={x f (1),x f (2),x f (3),…,x f (n)};
The processor filters X by using a fourth-order Butterworth filter with the bandwidth of 0.5Hz-9Hz, filters out signal data of a non-heart rate frequency band, eliminates noise and obtains a BCG signal X after the filtering treatment f ;
Step 2, extracting X by using a nonlinear differential enhancement method f Non-linear differential enhancement feature set X of (2) fd Wherein X is fd ={x fd (1),x fd (2),x fd (3),…,x fd (n)};
Step 3, for X f Extracting a wide time scale feature set X fw :
x fw (i)=x f (i-t w );
X fw ={x fw (1),x fw (2),x fw (3),…,x fw (n)};
Wherein x is fw (i) Is X fw Elements of (a) and (b);
step 4, for X f Extracting narrow time scale feature set X fs :
x fs (i)=x f (i-t s );
X fs ={x fs (1),x fs (2),x fs (3),…,x fs (n)};
Wherein x is fs (i) Is X fs Elements of (a) and (b);
step 5, using two fourth order Butterworth low pass filters to make X f Filtering respectively, and extracting X with cut-off frequencies of 2Hz and 4Hz of two fourth-order Butterworth low-pass filters f Feature set X of two frequency bands fl And X fh ;
Wherein X is fl ={x fl (1),x fl (2),x fl (3),…,x fl (n)};
X fh ={x fh (1),x fh (2),x fh (3),…,x fh (n)};
x fl (i) Is X fl Element x of (a) fh (i)X fh Elements of (a) and (b);
step 6, calculating an autocorrelation function F corresponding to the feature space F (i) based on the multidimensional features r (i);
Step 7, find F r (i) Determining heart rates respectively corresponding to the multidimensional features by using the maximum peak value in the peak set;
and 8, placing H into different heart rate intervals, selecting one or two heart rate intervals with the largest amount of heart rate data, calculating the average value of all heart rate data in all selected heart rate intervals, and taking the average value as the obtained final heart rate.
2. The BCG signal heart rate extraction method based on multi-dimensional features of claim 1, wherein step 2 comprises the steps of:
step 2-1, for X f Performing first-order median difference processing to obtain median difference characteristics x d (i):
x d (i1)=(x f (i1-1)-x f (i1+1))/2,i1=2,3,…,n-1;
x d (1)=x f (1),x d (n)=x f (n),i=1,2,3,…,n;
Step 2-2, for x d (i) Scaling is performed:
find x d (1),x d (2),x d (3),…,x d Maximum value x in (n) dmax By usingCalculating x d average ;
Using the formula alpha x d (i)/(x dmax ×x d average ) For x d (i) Scaling, wherein alpha is a scaling parameter, and alpha is more than or equal to 1 and less than or equal to 8;
step 2-3, performing feature amplification by using a nonlinear mapping function to obtain nonlinear differential enhancement features X fd :
X fd ={x fd (1),x fd (2),x fd (3),…,x fd (n)}。
3. The BCG signal heart rate extraction method based on multi-dimensional features of claim 2, wherein step 6 comprises the steps of:
the autocorrelation function F corresponding to the feature space F (i) based on the multidimensional feature is calculated by using the following formula r (i):
Wherein F (i) = { x fd (i),x fw (i),x fs (i),x fl (i),x fh (i)};
F r (i)={x rfd (i),x rfw (i),x rfs (i),x rfl (i),x rfh (i)};
k=0,1,2,…,n-1;
4. The BCG signal heart rate extraction method based on multi-dimensional features of claim 3, wherein step 7 comprises the steps of:
F r (i) Each element of the set comprises n data, and the n data of each element are subjected to normalization processing to obtain five sets Fr1, fr2, fr3, fr4 and Fr5 with the length of n: fr 1= { Fr1 (1), fr1 (2), fr1 (3), …, fr1 (n) };
Fr2={Fr2(1),Fr2(2),Fr2(3),…,Fr2(n)};
Fr3={Fr3(1),Fr3(2),Fr3(3),…,Fr3(n)};
Fr4={Fr4(1),Fr4(2),Fr4(3),…,Fr4(n)};
Fr5={Fr5(1),Fr5(2),Fr5(3),…,Fr5(n)};
the sets P1, P2, P3, P4 and P5 of all the peak points of Fr1, fr2, fr3, fr4, fr5 are determined respectively using the following formulas:
p1= { i1|fr1 (i 1) > Fr1 (i 1-1) and Fr1 (i 1) > Fr1 (i 1+1) };
p2= { i1|fr2 (i 1) > Fr2 (i 1-1) and Fr2 (i 1) > Fr2 (i1+1) };
p3= { i1|fr3 (i 1) > Fr3 (i 1-1) and Fr3 (i 1) > Fr3 (i1+1) };
p4= { i1|fr4 (i 1) > Fr4 (i 1-1) and Fr4 (i 1) > Fr4 (i1+1) };
p5= { i1|fr5 (i 1) > Fr5 (i 1-1) and Fr5 (i 1) > Fr5 (i1+1) };
let j be the number of the set of all peak points, j=1, 2,3,4,5; since the peak set Pj of the autocorrelation function represents a numberAccording to different time domain components, the maximum peak value Pj in Pj max Instantaneous heart rate H, which is a feature corresponding to Pj j The processor calculates and obtains an instantaneous heart rate set H:
H={H 1 ,H 2 ,H 3 ,H 4 ,H 5 store H into memory.
5. The multi-dimensional feature based BCG signal heart rate extraction method of claim 4, wherein step 8 comprises the steps of:
setting heart rate intervals which are sequentially arranged from low to high: [30, 35) bpm, [35, 40) bpm, [40, 45) bpm, [45, 50) bpm, [50, 55) bpm, …, [165, 170) bpm, [170, 175) bpm, [175,180] bpm;
the processor will H in H 1 ,H 2 ,H 3 ,H 4 ,H 5 And respectively classifying the heart rate intervals with the values into different heart rate intervals, selecting one or two heart rate intervals with the largest quantity of heart rate data, calculating the average value of all heart rate data in all selected intervals, and taking the average value as the obtained final heart rate.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311010661.8A CN117033906A (en) | 2023-08-11 | 2023-08-11 | BCG signal heart rate extraction method based on multidimensional features |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311010661.8A CN117033906A (en) | 2023-08-11 | 2023-08-11 | BCG signal heart rate extraction method based on multidimensional features |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117033906A true CN117033906A (en) | 2023-11-10 |
Family
ID=88631258
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311010661.8A Pending CN117033906A (en) | 2023-08-11 | 2023-08-11 | BCG signal heart rate extraction method based on multidimensional features |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117033906A (en) |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080194975A1 (en) * | 2007-02-08 | 2008-08-14 | Heart Force Medical Inc. | Monitoring physiological condition and detecting abnormalities |
US20180279961A1 (en) * | 2014-11-03 | 2018-10-04 | Agency For Science, Technology And Research | Multi-channel ballistocardiography with cepstrum smoothing and quality-based dynamic channel selection |
CN110384490A (en) * | 2019-07-29 | 2019-10-29 | 杭州埃因霍温科技有限公司 | BCG signal heart rate extracting method based on phase space |
CN114027813A (en) * | 2021-10-25 | 2022-02-11 | 深圳市麦格米特控制技术有限公司 | Heart rate extraction method, device, equipment and medium |
CN114886416A (en) * | 2022-05-07 | 2022-08-12 | 重庆邮电大学 | Heart rate and cardiac cycle detection method based on ballistocardiogram signals |
CN115251875A (en) * | 2022-06-23 | 2022-11-01 | 成都乐享智家科技有限责任公司 | Multichannel BCG signal detection and analysis method and heart rate detection method and device |
CN115299911A (en) * | 2022-08-08 | 2022-11-08 | 济南中科泛在智能计算研究院 | Heart rate extraction method and system based on cardiac shock signal |
CN116211287A (en) * | 2022-12-30 | 2023-06-06 | 广州中科新知科技有限公司 | Physical sign information monitoring method, device and system based on ballistocardiogram |
KR20230081154A (en) * | 2021-11-30 | 2023-06-07 | (주)허니냅스 | Data processing apparatus for detecting heart rate interval and operating method of the same |
CN116369888A (en) * | 2023-03-16 | 2023-07-04 | 武汉理工大学 | Non-contact heart rate variability data acquisition method and device |
-
2023
- 2023-08-11 CN CN202311010661.8A patent/CN117033906A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080194975A1 (en) * | 2007-02-08 | 2008-08-14 | Heart Force Medical Inc. | Monitoring physiological condition and detecting abnormalities |
US20180279961A1 (en) * | 2014-11-03 | 2018-10-04 | Agency For Science, Technology And Research | Multi-channel ballistocardiography with cepstrum smoothing and quality-based dynamic channel selection |
CN110384490A (en) * | 2019-07-29 | 2019-10-29 | 杭州埃因霍温科技有限公司 | BCG signal heart rate extracting method based on phase space |
CN114027813A (en) * | 2021-10-25 | 2022-02-11 | 深圳市麦格米特控制技术有限公司 | Heart rate extraction method, device, equipment and medium |
KR20230081154A (en) * | 2021-11-30 | 2023-06-07 | (주)허니냅스 | Data processing apparatus for detecting heart rate interval and operating method of the same |
CN114886416A (en) * | 2022-05-07 | 2022-08-12 | 重庆邮电大学 | Heart rate and cardiac cycle detection method based on ballistocardiogram signals |
CN115251875A (en) * | 2022-06-23 | 2022-11-01 | 成都乐享智家科技有限责任公司 | Multichannel BCG signal detection and analysis method and heart rate detection method and device |
CN115299911A (en) * | 2022-08-08 | 2022-11-08 | 济南中科泛在智能计算研究院 | Heart rate extraction method and system based on cardiac shock signal |
CN116211287A (en) * | 2022-12-30 | 2023-06-06 | 广州中科新知科技有限公司 | Physical sign information monitoring method, device and system based on ballistocardiogram |
CN116369888A (en) * | 2023-03-16 | 2023-07-04 | 武汉理工大学 | Non-contact heart rate variability data acquisition method and device |
Non-Patent Citations (1)
Title |
---|
田浩辰;赵海文;郭士杰;刘今越;王旭之;: "基于振动加速度的无束缚心冲击检测与心率提取方法研究", 生物医学工程学杂志, no. 02, 6 March 2019 (2019-03-06), pages 1001 - 5515 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110664390B (en) | Heart rate monitoring system and method based on wrist strap type PPG and deep learning | |
CN116705337B (en) | Health data acquisition and intelligent analysis method | |
CN106108889A (en) | Electrocardiogram classification method based on degree of depth learning algorithm | |
CN109998522A (en) | A kind of exercise load detection device and method based on wearable dynamic electrocardiogram | |
CN103168306A (en) | Method for determining a feature of the circadian rhythm of a subject | |
CN106388825B (en) | Respiration rate extraction method and device | |
CN111696575A (en) | Low ventilation and apnea detection and identification system based on hybrid neural network model | |
CN105913066B (en) | A kind of digitlization lungs sound feature dimension reduction method based on Method Using Relevance Vector Machine | |
WO2018149147A1 (en) | Method and apparatus for extracting respiration rate | |
CN107403407A (en) | A kind of breathing tracking based on thermal imaging | |
CN109044348A (en) | Atrial fibrillation detection apparatus and storage medium | |
CN109222964A (en) | Atrial fibrillation detection apparatus and storage medium | |
CN110151165A (en) | A kind of monocardiogram classification method based on Nonlinear Dynamical Characteristics | |
CN115363586A (en) | Psychological stress grade assessment system and method based on pulse wave signals | |
CN114582503A (en) | Blood pressure monitoring method based on multiple linear regression and correlation weighting selection algorithm | |
CN117814805B (en) | Intelligent processing method for data of clinical care equipment | |
CN110384490A (en) | BCG signal heart rate extracting method based on phase space | |
CN117033906A (en) | BCG signal heart rate extraction method based on multidimensional features | |
CN114052741A (en) | Electrocardiogram recognition method based on multi-scale autoregressive model | |
CN106388824B (en) | Respiration rate extraction method and device | |
CN106725487A (en) | Respiration rate extraction method and device | |
CN116439690A (en) | Ballistocardiogram signal extraction method, ballistocardiogram signal extraction device, ballistocardiogram signal extraction equipment and storage medium | |
Ding-Fei et al. | Study of feature extraction based on autoregressive modeling in EGG automatic diagnosis | |
CN114246591B (en) | Intelligent auxiliary myocardial infarction verification method and system based on knowledge graph | |
CN114098721B (en) | Ballistocardiogram signal extraction method, ballistocardiogram signal extraction device and ballistocardiogram signal extraction equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |