CN114767127A - Method and system for processing ballistocardiogram signal - Google Patents
Method and system for processing ballistocardiogram signal Download PDFInfo
- Publication number
- CN114767127A CN114767127A CN202210705321.6A CN202210705321A CN114767127A CN 114767127 A CN114767127 A CN 114767127A CN 202210705321 A CN202210705321 A CN 202210705321A CN 114767127 A CN114767127 A CN 114767127A
- Authority
- CN
- China
- Prior art keywords
- signal
- bcg
- peak
- init
- bcg signal
- 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.)
- Granted
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
-
- 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
Abstract
The invention discloses a method and a system for processing ballistocardiogram signals, wherein the method comprises the following steps: s1, carrying out BCG signal acquisition, and storing the acquired BCG signal into a cache memory to obtain a BCG signal I; s2, preprocessing the BCG signal I; s3, obtaining a BCG signal envelope based on J peak characteristics from the BCG signal II through Hilbert transform; s4, obtaining a J peak coordinate sequence of the BCG signal II through the wave peak enveloped by the BCG signal; s5, constructing a multi-scale BCG template signal based on the average heart cycle; and S6, acquiring and intercepting a plurality of sample signals from the BCG signal I, and carrying out DTW algorithm operation on each sample signal and the multi-scale BCG template signal together to obtain a final J peak coordinate sequence of the BCG, so as to complete the processing of the ballistocardiogram signal and realize the positioning of the heartbeat cycle. The problems that BCG signals are easy to interfere, heartbeat cycle positioning is not accurate enough and the like are solved.
Description
Technical Field
The invention relates to the technical field of biomedical signal processing, in particular to a ballistocardiogram signal processing method and a ballistocardiogram signal processing system.
Background
A large number of clinical researches prove that the establishment of data of nighttime continuous basic vital signs, such as resting cardiac cycle, respiratory effort and the like, is the key for realizing the home management of patients suffering from cardiovascular and cerebrovascular diseases and respiratory chronic diseases. However, the conventional medical apparatus or wrist-worn device cannot be used at home for a long time due to strong constraint and poor experience when monitoring the night cardiac cycle rhythm and respiratory effort change process of the patient with the chronic disease, or cannot perform medical-level analysis and diagnosis due to insufficient accuracy.
In the periodic blood pumping process of the heart, a weak and dynamically changing mechanical signal, namely, a Ballistocardiogram original signal (hereinafter referred to as BCG), is generated in the direction of gravity of a human body, and the BCG can reflect the working states of the heart and the cardiovascular system. In 1939 Starr's study indicated that a normal BCG should be consistent with the heartbeat, be repetitive, and contain mainly H, I, J, K, L and M, N waves, of which H, I, J, K waves are the largest in amplitude, and the combination of which resembles the shape of the letter W, often referred to as the "W" combination wave, as shown in fig. 1.
BCG signals are very susceptible to various types of noise and motion artifacts. Therefore, the algorithm research for resisting interference is the key for improving the accuracy of BCG analysis. The currently common BCG heartbeat feature positioning method comprises the following steps: the heart beat interval is estimated by a pseudo cycle positioning method, a clustering self-adaptive template matching method and a signal entropy value analysis method. For example, the existing patent document CN201810104375.0 discloses a BCG signal analysis method and system for monitoring hypertensive patients, which is to construct a classifier based on class association rules to determine whether the BCG signal is a hypertensive signal; CN201310393391.3 discloses a method and system for extracting BCG signal features, which introduces the cumulative residual entropy algorithm into the feature extraction classification of BCG signals, performs feature extraction for different frequency bands of the BCG signals, can analyze and obtain which frequency band has a prominent contribution to disease classification, and classifies the analysis result, so as to achieve the purpose of accurately extracting useful features in the BCG signals to perform disease classification diagnosis. Although the methods can position the heartbeat interval, the anti-interference capability of the BCG signals obtained under different environments is not enough, and the calculation power required by the methods for calculating in the embedded device is not enough to quickly and effectively position the BCG heartbeat interval in practical application.
Disclosure of Invention
Aiming at the problems that BCG signals are easy to interfere, the heartbeat cycle positioning is not accurate enough and the like, the invention aims to provide a ballistocardiogram signal processing method and a ballistocardiogram signal processing system.
In order to realize the purpose, the invention adopts the following technical scheme:
a method of processing ballistocardiogram signals, comprising the steps of:
s1, carrying out BCG signal acquisition, and storing the acquired BCG signal into a cache memory to obtain a BCG signal I;
s2, preprocessing the BCG signal I, eliminating thermal noise, power frequency interference and high-frequency signal noise in the BCG signal through a low-pass filter, and eliminating baseline interference of low-frequency signals in the BCG signal through a morphological filter bank to obtain a BCG signal II;
s3, obtaining a BCG signal envelope based on J peak characteristics from the BCG signal II through Hilbert transform;
s4, obtaining a wave crest sequence by positioning the wave crest enveloped by the BCG signal, and calculating the maximum value coordinate in the wave crest sequence to obtain a J peak coordinate sequence of the BCG signal II;
s5, calculating an average heart cycle through the J peak coordinate sequence, and constructing a multi-scale BCG template signal based on the average heart cycle;
and S6, acquiring and intercepting a plurality of sample signals from the BCG signal I, and carrying out DTW algorithm operation on each sample signal and the multi-scale BCG template signal together to obtain a final J peak coordinate sequence of the BCG, so that the ballistocardiogram signal is processed, and the positioning of the heartbeat cycle is realized.
Further, the step S1 further includes:
s11, carrying out BCG signal acquisition, wherein the frequency of the BCG signal is 1000 Hz;
s12, establishing a buffer space with the capacity of 15S;
s13, storing the acquired BCG signal into a cache memory of the cache space to obtain a BCG signal I;
and S14, updating the BCG signal I in real time in an FIFO mode, wherein the BCG signal I is updated for 1S each time.
Further, the low-pass filter is a zero-phase bilinear Butterworth low-pass filter, and the cut-off frequency of the low-pass filter is 20 Hz.
Further, the step S2 further includes using two sets of morphological filter operations to extract the low frequency baseline interference signal, which includes the following specific steps:
s21, the first group of morphological filter obtains the signal MF through the first opening and then closing operation1The second group of morphological filters obtains the signal MF by a first-off and then-on operation2;
S22, signal MF1And MF (MF)2After combination, a low-frequency baseline signal MF (t) is obtained,
and S23, subtracting the baseline signal MF (t) from the low-pass filtered BCG signal I to obtain a BCG signal, namely the BCG signal II finally filtered of the low-frequency baseline interference.
Further, the step S3 further includes:
s31, performing Hilbert transform operation on the BCG signal II to obtain a BCG signal III with the phase displacement of 90 degrees, wherein the specific calculation formula is as follows:
whereinIs the BCG signal III, and is,is used as the BCG signal II and is used as the BCG signal,is a time constant, t is time, H represents Hilbert transform, s (t) represents BCG signal;
s32, calculating the amplitude of the BCG signal II to obtain the BCG signal envelope, wherein the specific formula is as follows:
Further, the step S4 further includes:
s41, calculating all peak coordinates appearing in the BCG signal envelope to form a peak sequence init _ peak (n) of the BCG signal envelope, wherein n is the number of coordinates in the peak sequence of the BCG signal envelope, n =1,2 … … i,
init_peak(n)=[init_peak0 , init_peak1 , ……,init_peaki];
s42, calculating the distance from each peak to a trough in the peak sequence init _ peak (n) to obtain a vertical distance sequence AP (n) of the BCG signal envelope;
s43, calculating the average amplitude m _ ap of the vertical distance sequence AP (n), and deleting the peak coordinates of which the vertical distance is smaller than the average amplitude m _ ap from the peak sequence init _ peak (n) to obtain a new peak sequence init _ peak2 (n);
init_peak2(n)=[init_peak20, init_peak21, ……,init_peak2i];
s44, designing tau =50ms, and taking a search interval as
[init_peak2i-τ,init_peak2i+τ]The width of the BCG signal II is traversed, the maximum value in the interval is calculated, and then the J peak coordinate sequence init _ Jpeak (n) of the current BCG signal is obtained through positioning,
init_Jpeak(n)=[init_Jpeak0, init_Jpeak1, ……, init_Jpeaki]。
further, the step S5 further includes:
s51, sequentially carrying out subtraction operation on adjacent coordinates in the J peak coordinate sequence to obtain a heartbeat interval sequence init _ interval (n);
s52, calculating by the heartbeat interval sequence init _ interval (n), and averaging the heartbeat intervals mean _ i;
s53, setting the length of the multi-scale template signal, wherein the specific calculation method comprises the following steps:
m _ Len = mean _ i × d, M _ Len is the length of the multi-scale template signal, mean _ i is the average heartbeat interval, d is a multiple of the average heartbeat interval mean _ i, d =1,2,. M, and the maximum length mean _ i × M is less than half of the signal length of the BCG signal ii.
Further, the step S6 further includes:
s61, setting a DTW operation length interval as l = [0,1 × Δ t, … …, (h-1) × Δ t, … … 1500], wherein Δ t =10ms, h is the number of sample signals, and h belongs to [0,150];
s62, adding init _ Jpeak on BCG signal S (t)0Is a starting coordinate, 150 sample signals S' are interceptedn,S`nThe samples are represented as:
S`n=S(v),v∈[init_Jpeak0,init_Jpeak0+h*Δt];
s63, for each sample S ″iCarrying out DTW algorithm operation with the template signal to obtain an accumulated distance D (h);
s64, calculating the coordinate mark of the minimum distance obtained by Min (D (h)) as the J peak coordinate of the BCG;
and S65, repeating the calculation, and traversing all init _ JPeak (n) to obtain JPeak (n) which is the J peak coordinate sequence of the finally obtained BCG.
A processing system of ballistocardiogram signals is used for realizing the processing method of the ballistocardiogram signals, and comprises a BCG signal acquisition unit, a preprocessing unit, a wave crest extraction and calculation unit, a multi-scale BCG template signal construction unit and a DTW algorithm operation unit;
the BCG signal acquisition unit is used for acquiring and updating BCG signals;
the preprocessing unit is used for eliminating thermal noise, power frequency interference, high-frequency signal noise and baseline interference of low-frequency signals in the BCG signals;
the wave crest extraction and calculation unit is used for obtaining a J peak coordinate sequence of the BCG signal;
the multi-scale BCG template signal construction unit is used for calculating an average heart cycle through the J peak coordinate sequence and constructing a multi-scale BCG template signal;
the DTW algorithm operation unit is used for obtaining a plurality of sample signals, and performing DTW algorithm operation on each sample signal and the multi-scale BCG template signal together to obtain a final J peak coordinate sequence of the BCG.
The invention has the beneficial effects that:
1. the BCG signal of the aliasing breathing baseline is effectively extracted by adopting a mathematical morphological filtering mode, the waveform characteristics of the BCG signal on a time domain are greatly reserved, and the signal-to-noise ratio of the BCG signal is improved; the signal envelope of the BCG signal is efficiently extracted by combining with the Hilbert transform operation, and the judgment of the heart rate characteristic range of the BCG signal is realized.
2. A multi-scale BCG signal template is constructed based on the pre-estimated heart rate characteristic range, the accuracy of positioning the J peak of the BCG signal is improved by combining a dynamic time warping matching algorithm, and the range of calculating the effective heart rate through the BCG signal is effectively expanded.
Drawings
FIG. 1 is a waveform diagram of characteristics of a BCG signal;
FIG. 2 is a flow chart of a BCG heart cycle positioning method provided by the present invention;
FIG. 3 is a schematic diagram of the method of the present invention for removing low frequency baseline interference signals using a morphological filter;
FIG. 4 is a flow chart of the method for removing low frequency baseline interference signals by the morphological filter of the present invention;
FIG. 5 is a schematic diagram of a ballistocardiogram signal processing system according to the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings and the detailed description below:
the invention provides a ballistocardiogram signal processing method, which comprises the following steps:
s1, carrying out BCG signal acquisition, and storing the acquired BCG signal into a cache memory to obtain a BCG signal I;
step S1 further includes:
s11, carrying out BCG signal acquisition, wherein the frequency of the BCG signal is 1000 Hz;
s12, establishing a buffer space with the capacity of 15S;
s13, storing the acquired BCG signal into a cache memory of the cache space to obtain a BCG signal I;
and S14, updating the BCG signal I in real time in an FIFO mode, and updating the BCG signal I for 1S each time.
FIFO: simply understood as first-in first-out, is described in actual use as: in a 15s signal buffer array, storing data of the first second into an array position 0, after the 15s signal is stored fully, removing a 1s signal when a 16s signal comes in, moving a 2s signal to the array position 0, and positioning the 16s signal at the array position of the original 15s signal; by the method, the real-time performance of signal processing can be effectively kept, and the use of the memory of the embedded system is saved.
S2: preprocessing and filtering the BCG signal I to eliminate signal power frequency interference; baseline interference of low frequency signals within 0.5HZ is removed by a morphological filter.
Preprocessing the BCG I signal in the cache memory, including eliminating thermal noise, power frequency interference and high-frequency signal noise in the BCG signal by using a zero-phase bilinear Butterworth low-pass filter with the cut-off frequency of 20Hz, separating out low-frequency baseline interference signals within 0.5Hz in the BCG signal by using a morphological filter combination, and finally obtaining the BCG signal with the high-frequency thermal noise and the low-frequency baseline interference removed, as shown in FIG. 3.
The original signal before preprocessing is formed by superposing two signals of aliasing respiratory motion and a heartbeat BCG signal. Preprocessing and filtering: the signal from the hardware can have power frequency interference and a high-frequency thermal noise signal, a power frequency signal of 50HZ is removed through power frequency filtering, and the main frequency of the BCG signal for cardiac shock is concentrated at 0-20Hz, so the main signal frequency for cardiac shock is extracted through a Butterworth low-pass filter, the cost of hardware filtering can be reduced through a software filtering mode, the time delay of IIR (infinite impulse response) filtering can be eliminated through bilinear zero-phase Butterworth filtering, and meanwhile, the filtered signal result is better.
The method uses two sets of morphological filter operations to extract low-frequency baseline interference signals MF (t), and as shown in figure 1, sets the signal length of a structural function g (m) of a morphological filter to be 200ms and a linear signal with a numerical value of zero on the basis of H, I, J, K characteristic width of effective BCG under 1KHz sampling; the first group of morphological filters obtains the signal MF by a first-to-open and then-to-close operation1And a second set of morphological filters for obtaining the signal MF by a first-off-first-on operation2Signal MF1And MF (MF)2And obtaining a low-frequency baseline signal MF (t) after combination, and then subtracting the baseline signal MF (t) from the low-pass filtered BCG signal s (t) to obtain the BCG signal which is finally filtered of the low-frequency baseline interference. Referring to fig. 4, the design flow of the morphological filter in the invention method is a design flow chart of removing the low-frequency baseline interference signal by the morphological filter. The low-frequency baseline interference signal refers to a respiration baseline, the generation source of the respiration baseline is that when a human body is measured, due to the respiratory motion of the human body, low-frequency signal interference of 0-1HZ can be generated, the time domain wave characteristic of the BCG signal can be effectively reserved by using the morphological filter, the peak characteristic points of H/I/J/K/L of the BCG signal can be extracted through the characteristic of the morphological filter, the main characteristic of the BCG signal is better reserved, and the positioning of subsequent heart rate calculation is more efficient.
S3: obtaining a BCG signal envelope based on J peak characteristics from the BCG signal II through Hilbert Transform (HT), wherein the specific calculation mode is as follows:
performing HT operation on the 15s BCG signal s (t) to obtain a 90-degree phase-shifted BCG signalThe hilbert transform calculation formula is as follows:
whereinIn order to be the BCG signal ii,is a time constant, t is time, H represents Hilbert transform, s (t) represents BCG signal;
calculating the amplitude value of the BCG signal according to the following formula to obtain a BCG signal envelope e (t):
wherein, the first and the second end of the pipe are connected with each other,denoted BCG signal iii, s (t) is BCG signal.
Namely, the BCG signal with a phase shift of 90 degrees can be obtained according to the formula (1)Combining the BCG signal s (t) with the HT transformed signal according to equation (2)The BCG signal envelope e (t) is obtained.
By calculating and extracting the signal envelope of the BCG, the heartbeat interval of the signals in the 15s cache can be estimated, an average heartbeat interval of the 15s signals is obtained, a reference range is provided for subsequent BCG signal J peak positioning, a basis is provided for self-adaptive identification of the heartbeat interval, and the peak value of the envelope signal can provide an auxiliary positioning effect when the subsequent BCG signal J peak extraction is carried out.
S4: the BCG signal envelope e (t) obtained in S3 has time sequence consistency with the BCG signal S (t), so the peak sequence init _ peak (n) obtained by locating the peak of the signal envelope e (t) calculates the coordinate of the maximum value on the BCG signal S (t) in the search interval of 50ms to obtain the J peak coordinate sequence init _ jpeak (n) of the BCG. The heartbeat cycle can be reflected by the spacing of the J-peak coordinates subsequently,
s4: the method for realizing the initial positioning of the J peak of the heartbeat characteristics of the BCG signal comprises the following steps:
s41, calculating all peak coordinates appearing in the signal e (t) in S3 to form a peak sequence init _ peak (n) of the BCG signal envelope, wherein n is the number of coordinates in the peak sequence of the BCG signal envelope, n =1,2 … … i,
init_peak(n)=[init_peak0 , init_peak1 , ……,init_peaki];
s42, calculating the distance from each peak to a trough in the init _ peak (n) to obtain a vertical distance sequence AP (n) of the BCG envelope;
s43, calculating the average amplitude m _ ap of the sequence AP (n), deleting the peak coordinates with the vertical distance smaller than m _ ap from the init _ peak (n) to obtain a new peak sequence init _ peak2(n), init _ peak2(n) = [ init _ peak2(n) =0, init_peak21, ……,init_peak2i];
S44,: designing tau =50ms, taking the search interval as the BCG signal s (t)
[init_peak2i-τ,init_peak2i+τ]And calculating the maximum value in the interval, thereby positioning and obtaining the J peak coordinate sequence init _ Jpeak (n) of the current BCG signal, init _ Jpeak (n) = [ init _ Jpeak = [ ]0, init_Jpeak1, ……, init_Jpeaki]。
S5: construction of multi-scale BCG template based on average heart cycle
and S51, performing adjacent coordinate subtraction operation on the J peak coordinate sequence obtained in the S4 to obtain a heartbeat interval sequence init _ interval (n).
S52, calculating to obtain an average heartbeat interval mean _ i; the average heartbeat interval refers to calculating an average value of intervals in an init _ interval (n) sequence to obtain an average heartbeat interval;
s53, setting the length of the multi-scale template signal as M _ Len = mean _ i x d according to the multiple of mean _ i, wherein M _ Len is the length of the multi-scale template signal, mean _ i is an average heartbeat interval, d is the multiple of the average heartbeat interval mean _ i, d =1,2,. M, and the maximum length mean _ i x M is less than half of the length of the BCG signal II;
the longer the maximum length of M _ Len, mean _ i, the less interval data that can be calculated at the end, meaning that a portion of the data result is lost, and the longer the length of M _ Len, the more computation is required. Therefore, in the present invention, the maximum length of M _ Len is defined as half of the signal length.
S54, in the BCG signal S (t) according to the peak sequence init _ Jpeak (n), n =0,1,2,. K-1 according to the J peak sequence init _ Jpeak (n), with each J peak coordinate init _ JpeakiAs a center, the signal length with the length of t1=200ms is intercepted to the left, and the signal length with the length of t2=300ms is intercepted to the right, so that K BCG sample signals mt are obtainedi(t),t=[init_peaki-t1,init_peaki-t2],i=0,1,2,……K-1;
S55, according to the length M of the multi-scale template, a multi-scale template signal can be constructedComprises the following steps:
s6: and carrying out heartbeat cycle positioning through dynamic time warping.
The DTW algorithm is used in the invention, because the BCG signal is a pseudo-periodic signal, the peak value positioning can be carried out on the pseudo-periodic signal through the characteristics of the DTW algorithm, the distance between the peak value of the current period and the peak value of the next period is different, but the characteristics of the waveform are similar, and the peak value positioning can be carried out on the signals in a more effective self-adaptive interval through the DTW algorithm.
Setting the DTW operation length interval as l = [0,1 × Δ t, … …, (h-1) × Δ t, … … 1500 according to the algorithm thought of Dynamic Time Warping (DTW)]Δ t =10ms, h is the number of sample signals, h ∈ [0,150]]Init _ Jpeak on BCG signal s (t)0Peak of (2)Truncating 150 sample signals S' with values of start coordinatesn,S`nThe samples are expressed as formula (3), and for each sample S ″nAnd template signalAnd (3) carrying out DTW algorithm operation to obtain an accumulated distance D (h), calculating a coordinate mark of the minimum distance obtained by Min (D (h)) as a J peak coordinate of the BCG, repeating the calculation, and traversing all init _ Jpeak (n) obtained Jpeak (n) to obtain a J peak coordinate sequence of the BCG finally.
S`n=S(v),v∈[init_Jpeak0,init_Jpeak0+h*Δt](3)
In the practical application process, the steps S1-S6 are continuously and circularly repeated to process the BCG signals, so that the BCG heartbeat interval can be quickly and effectively positioned.
As shown in fig. 5, a processing system of ballistocardiogram signals, which is used for implementing the processing method of the ballistocardiogram signals, comprises a BCG signal acquisition unit, a preprocessing unit, a peak extraction computing unit, a multi-scale BCG template signal construction unit and a DTW algorithm operation unit;
the BCG signal acquisition unit is used for acquiring and updating BCG signals;
the preprocessing unit is used for eliminating thermal noise and power frequency interference in the BCG signal and baseline interference of high-frequency signal noise and low-frequency signal;
the wave crest extraction and calculation unit is used for obtaining a J peak coordinate sequence of the BCG signal;
the multi-scale BCG template signal construction unit is used for calculating an average heart cycle through the J peak coordinate sequence and constructing a multi-scale BCG template signal;
the DTW algorithm operation unit is used for obtaining a plurality of sample signals, and carrying out DTW algorithm operation on each sample signal and the multi-scale BCG template signal together to obtain a final J peak coordinate sequence of the BCG.
Various other modifications and changes may occur to those skilled in the art based on the foregoing teachings and concepts, and all such modifications and changes are intended to be included within the scope of the appended claims.
Claims (9)
1. A method of processing ballistocardiogram signals, comprising the steps of:
s1, carrying out BCG signal acquisition, and storing the acquired BCG signal into a cache memory to obtain a BCG signal I;
s2, preprocessing the BCG signal I, eliminating thermal noise, power frequency interference and high-frequency signal noise in the BCG signal through a low-pass filter, and eliminating baseline interference of low-frequency signals in the BCG signal through a morphological filter bank to obtain a BCG signal II;
s3, obtaining a BCG signal envelope based on J peak characteristics from the BCG signal II through Hilbert transform;
s4, obtaining a wave crest sequence by positioning the wave crest enveloped by the BCG signal, and calculating the maximum value coordinate in the wave crest sequence to obtain a J peak coordinate sequence of the BCG signal II;
s5, calculating an average heart cycle through the J peak coordinate sequence, and constructing a multi-scale BCG template signal based on the average heart cycle;
and S6, acquiring and intercepting a plurality of sample signals from the BCG signal I, and carrying out DTW algorithm operation on each sample signal and the multi-scale BCG template signal together to obtain a final J-peak coordinate sequence of the BCG so as to complete the processing of the ballistocardiogram signal.
2. The method for processing ballistocardiogram signals according to claim 1, wherein the step S1 further includes:
s11, carrying out BCG signal acquisition, wherein the frequency of the BCG signal is 1000 Hz;
s12, establishing a buffer space with the capacity of 15S;
s13, storing the acquired BCG signal into a cache memory of the cache space to obtain a BCG signal I;
and S14, updating the BCG signal I in real time in an FIFO mode, and updating the BCG signal I for 1S each time.
3. A ballistocardiogram signal processing method according to claim 1, wherein the low-pass filter is a zero-phase bilinear Butterworth low-pass filter, and the cut-off frequency of the low-pass filter is 20 Hz.
4. The method for processing ballistocardiogram signals according to claim 1, wherein the step S2 further includes using two sets of morphological filter operations to extract the low frequency baseline interference signal, the specific steps are as follows:
s21, the first group of morphological filter obtains the signal MF through the first opening and then closing operation1The second set of morphological filters obtain the signal MF by a first-off and then-on operation2;
S22, signal MF1And MF (MF)2Obtaining a low-frequency baseline signal MF (t) after combination;
and S23, subtracting the baseline signal MF (t) from the low-pass filtered BCG signal I to obtain a BCG signal, and finally filtering the low-frequency baseline interference BCG signal II.
5. The method for processing ballistocardiogram signals according to claim 1, wherein the step S3 further includes:
s31, performing Hilbert transform operation on the BCG signal II to obtain a BCG signal III with the phase displacement of 90 degrees, wherein the specific calculation formula is as follows:
whereinIs the BCG signal III, and the BCG signal III,in order to be the BCG signal ii,is a time constant, t is time, H represents Hilbert transform, s (t) represents BCG signal;
s32, calculating the amplitude of the BCG signal II to obtain the BCG signal envelope, wherein the specific formula is as follows:
6. The method for processing ballistocardiogram signals according to claim 1, wherein the step S4 further includes:
s41, calculating all peak coordinates appearing in the BCG signal envelope to form a peak sequence init _ peak (n) of the BCG signal envelope, wherein n is the number of coordinates in the peak sequence of the BCG signal envelope, n =1,2 … … i,
init_peak(n)=[init_peak0 , init_peak1 , ……,init_peaki];
s42, calculating the distance from each peak to a trough in the peak sequence init _ peak (n) to obtain a vertical distance sequence AP (n) of the BCG signal envelope;
s43, calculating the average amplitude m _ ap of the vertical distance sequence AP (n), deleting the peak coordinates of which the vertical distances are smaller than the average amplitude m _ ap from the peak sequence init _ peak (n) to obtain a new peak sequence init _ peak2(n),
init_peak2(n)=[init_peak20, init_peak21, ……,init_peak2i];
s44, designing tau =50ms, and taking a search interval as
[init_peak2i-τ,init_peak2i+τ]The width of the BCG signal II is traversed, the maximum value in the interval is calculated, and the position is obtainedThe J-peak coordinate sequence init _ jeak (n) of the front BCG signal,
init_Jpeak(n)=[init_Jpeak0, init_Jpeak1, ……, init_Jpeaki]。
7. the method for processing ballistocardiogram signals according to claim 6, wherein the step S5 further includes:
s51, sequentially carrying out subtraction operation on adjacent coordinates in the J peak coordinate sequence to obtain a heartbeat interval sequence init _ interval (n);
s52, calculating an average heartbeat interval mean _ i through the heartbeat interval sequence init _ interval (n);
s53, setting the length of the multi-scale template signal, wherein the specific calculation method comprises the following steps:
m _ Len = mean _ i × d, M _ Len is the length of the multi-scale template signal, mean _ i is the average heartbeat interval, d is a multiple of the average heartbeat interval mean _ i, d =1,2,. M, and the maximum length mean _ i × M is less than half of the length of the BCG signal ii.
8. The method for processing ballistocardiogram signals according to claim 7, wherein the step S6 further includes:
s61, setting DTW operation length interval as
l = [0,1 × Δ t, … …, (h-1) × Δ t, … … 1500], Δ t =10ms, h is sample signal
The number of numbers h belongs to [0,150];
s62, adding init _ Jpeak on BCG signal S (t)0Is a starting coordinate, truncating 150 sample signals S ″n,S`nThe samples are represented as:
S`n=S(v),v∈[init_Jpeak0,init_Jpeak0+h*Δt];
s63, for each sample S ″iCarrying out DTW algorithm operation with the template signal to obtain an accumulated distance D (h);
s64, calculating the coordinate mark of the minimum distance obtained by Min (D (h)) as the J peak coordinate of the BCG;
and S65, repeating the calculation, and traversing all init _ JPeak (n) to obtain JPeak (n) which is the J peak coordinate sequence of the finally obtained BCG.
9. A processing system of ballistocardiogram signals is used for realizing the processing method of the ballistocardiogram signals, and is characterized by comprising a BCG signal acquisition unit, a preprocessing unit, a wave crest extraction and calculation unit, a multi-scale BCG template signal construction unit and a DTW algorithm operation unit;
the BCG signal acquisition unit is used for acquiring and updating BCG signals;
the preprocessing unit is used for eliminating thermal noise and power frequency interference in the BCG signal and baseline interference of high-frequency signal noise and low-frequency signal;
the wave crest extraction and calculation unit is used for obtaining a J peak coordinate sequence of the BCG signal;
the multi-scale BCG template signal construction unit is used for calculating an average heart cycle through the J peak coordinate sequence and constructing a multi-scale BCG template signal;
the DTW algorithm operation unit is used for obtaining a plurality of sample signals, and carrying out DTW algorithm operation on each sample signal and the multi-scale BCG template signal together to obtain a final J peak coordinate sequence of the BCG.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210705321.6A CN114767127B (en) | 2022-06-21 | 2022-06-21 | Method and system for processing ballistocardiogram signal |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210705321.6A CN114767127B (en) | 2022-06-21 | 2022-06-21 | Method and system for processing ballistocardiogram signal |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114767127A true CN114767127A (en) | 2022-07-22 |
CN114767127B CN114767127B (en) | 2022-09-30 |
Family
ID=82421400
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210705321.6A Active CN114767127B (en) | 2022-06-21 | 2022-06-21 | Method and system for processing ballistocardiogram signal |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114767127B (en) |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107233086A (en) * | 2016-03-28 | 2017-10-10 | 三星电子株式会社 | The method and apparatus that heart rate and respiratory rate are estimated are carried out using low power sensor |
US20180000426A1 (en) * | 2016-06-29 | 2018-01-04 | Samsung Electronics Co., Ltd. | System and Method for Providing a Real-Time Signal Segmentation and Fiducial Points Alignment Framework |
CN107569226A (en) * | 2017-09-27 | 2018-01-12 | 广州中科新知科技有限公司 | HRV method and application is obtained based on piezoelectric sensing |
US20180128861A1 (en) * | 2016-11-07 | 2018-05-10 | Samsung Electronics Co., Ltd. | Representative waveform providing apparatus and method |
US20190166030A1 (en) * | 2012-12-05 | 2019-05-30 | Origin Wireless, Inc. | Method, apparatus, server and system for vital sign detection and monitoring |
EP3492003A1 (en) * | 2017-12-01 | 2019-06-05 | Samsung Electronics Co., Ltd. | Bio-signal quality assessment apparatus and bio-signal quality assessment method |
US20190167200A1 (en) * | 2017-12-01 | 2019-06-06 | Samsung Electronics Co., Ltd. | Bio-signal quality assessment apparatus and bio-signal quality assessment method |
US20190313947A1 (en) * | 2018-04-17 | 2019-10-17 | Samsung Electronics Co., Ltd. | Systematic apparatus for motion sensor and optical sensor based cardiac arrhythmia triage |
CN110420019A (en) * | 2019-07-29 | 2019-11-08 | 西安电子科技大学 | A kind of depth recurrence heart rate estimation method of ballistocardiography signal |
CN111035367A (en) * | 2019-12-31 | 2020-04-21 | 华南师范大学 | Signal detection method and system for judging sleep apnea |
CN111091116A (en) * | 2019-12-31 | 2020-05-01 | 华南师范大学 | Signal processing method and system for judging arrhythmia |
CN112507784A (en) * | 2020-10-30 | 2021-03-16 | 华南师范大学 | Validity detection method for ballistocardiogram time sequence signal |
CN112869733A (en) * | 2021-01-08 | 2021-06-01 | 广州中科新知科技有限公司 | Real-time heart beat interval measuring and calculating method for ballistocardiogram |
CN114376564A (en) * | 2021-12-29 | 2022-04-22 | 华南理工大学 | Sleep staging method, system, device and medium based on cardiac shock signal |
US20220167875A1 (en) * | 2019-03-28 | 2022-06-02 | Paris Sciences Et Lettres | Ballistocardiography device and method |
-
2022
- 2022-06-21 CN CN202210705321.6A patent/CN114767127B/en active Active
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190166030A1 (en) * | 2012-12-05 | 2019-05-30 | Origin Wireless, Inc. | Method, apparatus, server and system for vital sign detection and monitoring |
CN107233086A (en) * | 2016-03-28 | 2017-10-10 | 三星电子株式会社 | The method and apparatus that heart rate and respiratory rate are estimated are carried out using low power sensor |
US20180000426A1 (en) * | 2016-06-29 | 2018-01-04 | Samsung Electronics Co., Ltd. | System and Method for Providing a Real-Time Signal Segmentation and Fiducial Points Alignment Framework |
US20180128861A1 (en) * | 2016-11-07 | 2018-05-10 | Samsung Electronics Co., Ltd. | Representative waveform providing apparatus and method |
CN107569226A (en) * | 2017-09-27 | 2018-01-12 | 广州中科新知科技有限公司 | HRV method and application is obtained based on piezoelectric sensing |
EP3492003A1 (en) * | 2017-12-01 | 2019-06-05 | Samsung Electronics Co., Ltd. | Bio-signal quality assessment apparatus and bio-signal quality assessment method |
US20190167200A1 (en) * | 2017-12-01 | 2019-06-06 | Samsung Electronics Co., Ltd. | Bio-signal quality assessment apparatus and bio-signal quality assessment method |
CN109864704A (en) * | 2017-12-01 | 2019-06-11 | 三星电子株式会社 | Bio signal quality evaluation equipment and bio signal quality evaluating method |
US20190313947A1 (en) * | 2018-04-17 | 2019-10-17 | Samsung Electronics Co., Ltd. | Systematic apparatus for motion sensor and optical sensor based cardiac arrhythmia triage |
US20220167875A1 (en) * | 2019-03-28 | 2022-06-02 | Paris Sciences Et Lettres | Ballistocardiography device and method |
CN110420019A (en) * | 2019-07-29 | 2019-11-08 | 西安电子科技大学 | A kind of depth recurrence heart rate estimation method of ballistocardiography signal |
CN111035367A (en) * | 2019-12-31 | 2020-04-21 | 华南师范大学 | Signal detection method and system for judging sleep apnea |
CN111091116A (en) * | 2019-12-31 | 2020-05-01 | 华南师范大学 | Signal processing method and system for judging arrhythmia |
CN112507784A (en) * | 2020-10-30 | 2021-03-16 | 华南师范大学 | Validity detection method for ballistocardiogram time sequence signal |
CN112869733A (en) * | 2021-01-08 | 2021-06-01 | 广州中科新知科技有限公司 | Real-time heart beat interval measuring and calculating method for ballistocardiogram |
CN114376564A (en) * | 2021-12-29 | 2022-04-22 | 华南理工大学 | Sleep staging method, system, device and medium based on cardiac shock signal |
Also Published As
Publication number | Publication date |
---|---|
CN114767127B (en) | 2022-09-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Merah et al. | R-peaks detection based on stationary wavelet transform | |
Park et al. | R peak detection method using wavelet transform and modified shannon energy envelope | |
CN109907752B (en) | Electrocardiogram diagnosis and monitoring system for removing motion artifact interference and electrocardio characteristic detection | |
WO2017148452A1 (en) | Electrocardiography signal waveform feature point extraction method and device | |
US9642543B2 (en) | Systems and methods for model-based non-contact physiological data acquisition | |
Kumar et al. | Investigating cardiac arrhythmia in ECG using random forest classification | |
Sasikala et al. | Robust r peak and qrs detection in electrocardiogram using wavelet transform | |
CN108272451B (en) | QRS wave identification method based on improved wavelet transformation | |
Sabherwal et al. | Automatic detection of the R peaks in single-lead ECG signal | |
EP3270775B1 (en) | Systems, apparatuses and methods for sensing fetal activity | |
Nabar et al. | GeM-REM: Generative model-driven resource efficient ecg monitoring in body sensor networks | |
Zhang et al. | QRS detection based on morphological filter and energy envelope for applications in body sensor networks | |
Zou et al. | An ultra-low power QRS complex detection algorithm based on down-sampling wavelet transform | |
Berwal et al. | Design of high performance QRS complex detector for wearable healthcare devices using biorthogonal spline wavelet transform | |
WO2017084546A1 (en) | Wearable device-based user interest information determination method, device and wearable device | |
CN111317499A (en) | Heart sound signal processing method based on wavelet technology | |
Mabrouki et al. | R peak detection in electrocardiogram signal based on a combination between empirical mode decomposition and Hilbert transform | |
Chin et al. | Bayesian real-time QRS complex detector for healthcare system | |
Tang et al. | ECG de-noising based on empirical mode decomposition | |
CN109077721A (en) | atrial fibrillation detection device and storage medium | |
CN114027804A (en) | Pulse condition diagnosis method, device and readable storage medium | |
Braojos et al. | Embedded real-time ECG delineation methods: A comparative evaluation | |
CN102302373B (en) | Method for detecting heart sound signal feature points based on multiplexing multi-resolution wavelet transformation | |
CN114767127B (en) | Method and system for processing ballistocardiogram signal | |
CN110420022B (en) | P wave detection method based on dual-density wavelet transform |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |