KR102028517B1 - Heart rate variability analysis device and method of heart rate variability detection using the same - Google Patents

Abstract

The present invention relates to an apparatus for analyzing heart rate variability and a method for detecting heart rate variability using the same. The apparatus for analyzing heart rate variability according to an embodiment of the present invention includes a converter, an R-peak detector, a frequency converter, and a heart rate variance analyzer. The converter converts the analog ECG signal into a digital ECG signal. The R-peak detector receives the digital electrocardiogram signal to detect the R-peak value, and generates an RR interval time series signal based on the R-peak value. The frequency converter includes a frequency shifter, a low pass filter, and a fast Fourier transform. The frequency shifter extracts a signal of a target frequency band from an RR interval time series signal to generate a target frequency signal. The heart rate variance analyzer generates a heart rate variance signal based on the target frequency signal. According to the embodiment of the present invention, the frequency domain of the heart rate variability may be analyzed precisely.

Description

HEART RATE VARIABILITY ANALYSIS DEVICE AND METHOD OF HEART RATE VARIABILITY DETECTION USING THE SAME}

The present invention relates to the detection and analysis of heart rate variability, and more particularly, to an apparatus for analyzing heart rate variability and a method for detecting heart rate variability using the same.

The autonomic nervous system is involved in the maintenance of life-sustaining activity and homeostasis by regulating organ function and metabolism and by balancing the changes in environmental and external environmental factors. The autonomic nervous system is associated with a variety of mental, physical, and stressful disorders. Therefore, environmental stress can be evaluated by analyzing the state of the autonomic nervous system. Heart rate variability analysis can quantitatively assess the state of the sympathetic and parasympathetic nerves.

Analysis methods of heart rate variability include a time domain analysis method and a frequency domain analysis method. The time domain analysis method is a method of analyzing heart rate variability based on time intervals of QRS waves. For example, time domain analysis methods include SDNN, SDANN, RMSSD, NN50, and pNN50. The time domain analysis method is easy to calculate, but has a disadvantage in that the activity of the sympathetic nervous system and the parasympathetic nervous system cannot be distinguished or quantified in the balance of the autonomic nervous system.

On the other hand, the frequency domain analysis method is a method of evaluating the characteristics of the frequency band of the heart rate variability to evaluate the power value of the corresponding band. The frequency domain analysis method can quantify and show the balance of the autonomic nervous system related to the sympathetic nervous system and the parasympathetic nervous system. Accordingly, there is a continuing need for an apparatus and method for analyzing heart rate variability for frequency domain analysis.

The present invention can provide a heart rate variability analysis apparatus that can accurately analyze the frequency region of the heart rate variability and a heart rate variance detection method using the same.

The apparatus for analyzing heart rate variability according to an embodiment of the present invention includes a converter, an R-peak detector, a frequency converter, and a heart rate variance analyzer.

The converter receives an analog ECG signal containing a QRS wave and converts it into a digital ECG signal. The converter generates a digital ECG signal sampled at the sampling frequency.

The R-peak detector receives the digital electrocardiogram signal to detect the R-peak value, and generates an RR interval time series signal based on the R-peak value.

The frequency converter generates a target frequency signal by extracting a signal of a target frequency band from the RR interval time series signal. The frequency converter includes a frequency shifter, a low pass filter, a decimator, a data inserter, and a fast Fourier transform.

The frequency shifter shifts the target frequency band of the RR interval time series signal to the baseband to generate a shift detection signal. The frequency shifter generates an I channel signal by multiplying the RR interval time series signal by the first sinusoidal signal, and generates a Q channel signal by multiplying the RR interval time series signal by the second sinusoidal signal. The first sinusoidal signal and the second sinusoidal signal have a phase difference of 90 degrees, and the frequencies of the first sinusoidal signal and the second sinusoidal signal are based on the center frequency of the target frequency band. The low pass filter receives the shift detection signal and passes the baseband shift detection signal. The decimator reduces the sampling rate of the shift detection signal. The data inserter inserts zero data into the shift detection signal.

The heart rate variance analyzer generates a heart rate variance signal for the target frequency band based on the target frequency signal. The heart rate variability signal may include a power spectrum signal of a target frequency band. The heart rate variability analyzer may calculate at least one of a power value of a high frequency band, a power value of a low frequency band, a power value of an ultra low frequency band, and a power ratio of a low frequency band to a high frequency band based on the power spectrum signal.

The apparatus for analyzing heart rate variability may further include a band selector configured to provide target frequency information corresponding to the target frequency band mode to the frequency converter. The target frequency band mode may include a high frequency band mode, a low frequency band mode, and an ultra low frequency band mode.

The heart rate variance detection method according to an embodiment of the present invention includes receiving an ECG signal, detecting an R-peak value, converting a frequency, and calculating and analyzing the heart rate variance signal. The frequency conversion step includes shifting the target frequency band of the RR interval time series signal to the base band, filtering with a low pass filter, reducing the sampling rate of the RR interval time series signal, and inserting zero data into the RR interval time series signal. Step, and a fast Fourier transform step.

The apparatus for analyzing heart rate variability according to an embodiment of the present invention and a method for detecting heart rate variability using the same may detect and accurately analyze heart rate variability in a target frequency band.

1 is a graph showing the waveform of an ECG signal.
2 is a graph showing a waveform of a heart rate variability signal.
3 is a block diagram of an apparatus for analyzing heart rate variability according to an embodiment of the present invention.
4 is a block diagram of a frequency converter according to an embodiment of the present invention.
5 to 8 are graphs for describing a process of detecting a signal of a target frequency band by the frequency converter of FIG. 4.
9 is a block diagram of an apparatus for analyzing heart rate variability according to another embodiment of the present invention.
10 is a flowchart illustrating a method of detecting heart rate variability according to an embodiment of the present invention.
11 is a flowchart illustrating a step of converting a frequency according to an embodiment of the present invention.

In the following, embodiments of the present invention are described clearly and in detail so that those skilled in the art can easily practice the present invention.

1 is a graph showing the waveform of an ECG signal.

Electrocardiogram (ECG) is a graphical representation of the electrical activity occurring in the myocardium as the heart beats. Referring to FIG. 1, an ECG signal includes a P wave, a Q wave, an R wave, an S wave, and a T wave. 1 shows P waves, Q waves, R waves, S waves, and T waves by one heartbeat, but these waveforms are formed repeatedly.

P waves are due to contraction of the atria. The spacing between the P and Q waves corresponds to the time when the electrical stimulation of the atria is conducted to the atrioventricular node. In general, the interval between the P wave and the Q wave is formed within 0.2 seconds. The Q wave is a downward wave where the electrical current of the heart decreases. The R wave is an upward wave in which the electrical current of the heart increases after the Q wave. Due to the R wave, the ECG has a maximum value, which is defined as an R-peak value. The S wave is a downward wave in which the electrical current of the heart decreases after the R wave. Q waves, R waves, and S waves are due to the depolarization process of the ventricular muscles. The T wave is due to the repolarization process, which recovers and relaxes after ventricular contraction due to depolarization. Q waves, R waves, and S waves are defined as QRS waves, which are caused by depolarization in which electrical stimulation originating from the nodule is conducted through the atrioventricular node to the ventricles and constricts the ventricles. Generally, QRS waves are formed between 0.06 seconds and 0.1 seconds.

The waveform of the ECG signal of FIG. 1 illustrates the electrical activity of the heart over time. Heart rate variability analysis apparatus of the present invention detects the R-peak value of the waveform of the ECG signal, frequency conversion of the waveform of the heart rate tachogram showing the interval between the R- peak values by frequency conversion of the heart rate variability Analyze areas. In detail, the frequency conversion and heart rate variance signal generation process of the ECG signal will be described later.

2 is a graph showing a waveform of a heart rate variability signal.

Heart rate variability (HRV) means variability between heartbeats and heart rate variability, and the amount of change in the cycle may be detected by measuring the instantaneous period of the heartbeat. Heart rate variability may reflect the interaction of sympathetic and parasympathetic nerves.

Referring to FIG. 2, a power spectral density (PSD) for a frequency domain is illustrated as a waveform of a heart rate variability signal. The interval between heartbeats, such as the interval between R-peak values, is measured in time units, which time intervals generate a signal for frequency conversion. In power spectrum calculation, the unit of signal is squared. Since heart rate variance analysis calculates the power spectrum based on the interval between R-peak values obtained from time measurements, the unit of the vertical axis is defined in msec ² units.

The waveform of the heart rate variability signal is divided into a high frequency band (HF), a low frequency band (LF), and an ultra low frequency band (VLF). Power values of each of the high frequency band, the low frequency band, and the ultra low frequency band may be used to determine the activity of the sympathetic nervous system or the parasympathetic nervous system.

The high frequency band HF is defined as a frequency band between 0.15 and 0.4 Hz. The high frequency band (HF) can be used as an indicator of the activity of the parasympathetic nervous system. The high frequency band (HF) is associated with the breathing cycle. The high frequency band (HF) is related to the electrical stability of the heart. Therefore, by analyzing the power value of the high frequency band (HF) it is possible to determine the activity of the cardiopulmonary and electrical stability of the heart.

The low frequency band LF is defined as a frequency band between 0.04 and 0.15 Hz. The low frequency band LF is associated with changes in heart rate due to seizure reflex or blood pressure control. Low frequency band (LF) can be used as an indicator of the activity of the sympathetic nervous system. For example, when the activity of the sympathetic nervous system such as migraine patients is excessive, the fluctuation of the low frequency band (LF) is large. However, the present invention is not limited thereto, and the low frequency band LF may be used as an indicator of activity of the sympathetic nervous system and the parasympathetic nervous system.

The ultra low frequency band (VLF) is defined as a band between 0.003 and 0.04 Hz. Ultra low frequency band (VLF) can be used as an indicator of the activity of the sympathetic nervous system. For example, a sleep apnea patient may be stimulated by the sympathetic nerve during apnea to increase the power value of the ultra-low frequency band (VLF). In addition, the ultra low frequency band (VLF) may be used as an indicator for body temperature control, hormones, and vascular movement.

3 is a block diagram of an apparatus for analyzing heart rate variability according to an embodiment of the present invention.

Referring to FIG. 3, the heart rate variability analyzer 1000 includes a converter 100, an R-peak detector 200, a frequency converter 300, a heart rate variability analyzer 400, a band selector 500, and storage. The unit 600 and the display unit 700 are included.

The converter 100 receives an analog signal and converts it into a digital signal. The converter 100 receives the analog ECG signal ECGT and converts it into a digital ECG signal ECGn. The converter 100 is connected to an electrocardiogram sensor (not shown), and the electrocardiogram sensor that detects the electrocardiogram of the user provides the analog electrocardiogram signal ECGT to the converter 100. The analog ECG signal ECGT is converted into the digital ECG signal ECGn using the converter 100 for a specified measurement time. The digital electrocardiogram signal ECGn may be a discrete signal.

Converter 100 receives sampling signal fs. The sampling signal fs includes information about the sampling frequency of the ECG signal. The sampling frequency value may be formed between 200 and 500 Hz, but is not limited thereto. The converter 100 may determine the data amount of the converted digital electrocardiogram signal ECGn based on the sampling signal fs. If the frequency value of the sampling signal fs is large, the digital electrocardiogram signal ECGn with high stability may be generated. If the frequency value of the sampling signal fs is small, the digital ECG signal ECGn with less data may be generated. Can be.

The R-peak detector 200 receives the digital electrocardiogram signal ECGn from the converter 100. The R-peak detector 200 detects an R-peak value from the digital electrocardiogram signal ECGn. The R-peak detector 200 may include a QRS wave detection algorithm for detecting the R-peak value. For example, the R-peak detector 200 may include a Pan & Tompkins algorithm or a Hamilton & Tompkins algorithm. The R-peak detector 200 detects a plurality of R waves existing during the measurement time through a QRS wave detection algorithm, and sequentially detects R-peak values, which are peak values of the respective R waves.

The R-peak detector 200 generates the RR interval time series signal RPn based on the sequentially detected R-peak values. The R-peak detector 200 calculates an R-peak to R-peak interval (RRI) which is an interval between R-peak values sequentially detected as time passes. The R-peak detector 200 may interpolate the RR interval RRI to generate an RR interval time series signal RPn corresponding to the RRI tachogram. For example, based on continuous information of the RR interval (RRI), the R-peak detector 200 may have a sample density distribution such as a sample density distribution of the duration of the RR interval (RRI) or a sample density distribution of an adjacent RR interval (RRI) difference. The RR interval time series signal RPn may be generated through interpolation to convert the geometric pattern.

The frequency converter 300 receives the RR interval time series signal RPn from the R-peak detector 200. The frequency converter 300 generates the target frequency signal Tf by extracting the target frequency band ft of the RR interval time series signal RPn. The frequency converter 300 may convert the RR interval time series signal RPn into a frequency sequence signal using a fast Fourier transform (FFT).

The frequency converter 300 may extract a signal value corresponding to the target frequency band ft of the RR interval time series signal RPn before performing frequency conversion of the RR interval time series signal RPn. In order to extract a signal value corresponding to the target frequency band ft, the frequency converter 300 receives target frequency band information fti from the band selector 500 to be described later.

The target frequency band ft refers to a frequency band to be analyzed in the RR interval time series signal RPn. For example, if the subject of analysis is a high frequency band (HF), the target frequency band (ft) may include a band between 0.15 and 0.4 Hz. If the subject of the analysis is a low frequency band LF, the target frequency band ft may comprise a band between 0.04 and 0.15 Hz. If the subject of analysis is an ultra low frequency band (VLF), the target frequency band (ft) may include a band between 0.003 and 0.04 Hz. The frequency converter 300 generates the target frequency signal Tf based on the target frequency band information fti. The target frequency signal Tf may be a signal corresponding to the target frequency band ft in the heart rate variability signal for the frequency domain. In detail, the configuration of the frequency converter 300 and the process of generating the target frequency signal Tf will be described later.

The heart rate variability analyzer 400 receives the target frequency signal Tf from the frequency converter 300. The heart rate variability analyzer 400 analyzes the heart rate variability based on the target frequency signal Tf. The heart rate variance analyzer 400 generates a heart rate variance signal HRV for the target frequency band ft based on the target frequency signal Tf. The heart rate variance signal HRV may include a power spectrum signal PSD of a target frequency band ft. The heart rate variance signal HRV may be generated by calculating a power value of the target frequency signal Tf and calculating a power value.

For example, the heart rate variability signal HRV may be any one of a power value of the high frequency band HF, a power value of the low frequency band LF, or a power value of the ultra low frequency band VLF, depending on the target frequency band ft. It can be one. As described above, the power value of the high frequency band HF, the power value of the low frequency band LF, or the power value of the ultra low frequency band VLF may be used as an index for activity of the autonomic nervous system.

The heart rate variance signal HRV may include a power ratio LF / HF of the low frequency band LF to the high frequency band HF. If the power ratio (LF / HF) is high, it may mean that the sympathetic nerve is activated or the parasympathetic nerve is inhibited. The power ratio (LF / HF) allows quantification and analysis of the balance between the sympathetic and parasympathetic nervous systems. The heart rate variability signal HRV is not limited thereto and may include various analysis signals such as a power value of an ultra-low frequency band. Based on these various heart rate variability signals (HRV), heart rate variability analysis can be performed.

The band selector 500 provides the target frequency band information fti to the frequency converter 300. The target frequency band information fti may include information about the center frequency fc of the target frequency band ft or the target frequency band ft. The band selector 500 provides the target frequency band information fti corresponding to the target frequency band mode to the frequency converter 300 so that the heart rate variability analyzer 400 receives the heart rate variance signal of the target frequency band ft. To create it.

The target frequency band mode includes a plurality of frequency band modes. The target frequency band mode may include a high frequency band mode, a low frequency band mode, and an ultra low frequency band mode, and may include more band modes. The target frequency bands in each target frequency band mode may not overlap each other. In each target frequency band mode, the band selector 500 provides target frequency band information fti corresponding to each target frequency band ft to the frequency converter 300.

In the high frequency band mode, the band selector 500 provides the high frequency band information to the frequency converter 300. The frequency converter 300 receiving the high frequency band information generates the target frequency signal Tf for the high frequency band HF. The heart rate variance analyzer 400 generates a heart rate variance signal HRV for the high frequency band HF based on the target frequency signal Tf.

In the low frequency band mode, the band selector 500 provides the low frequency band information to the frequency converter 300. The frequency converter 300 receiving the low frequency band information generates the target frequency signal Tf for the low frequency band LF. The heart rate variance analyzer 400 generates a heart rate variance signal HRV for the low frequency band LF based on the target frequency signal Tf.

In the ultra low frequency band mode, the band selector 500 provides the ultra low frequency band information to the frequency converter 300. The frequency converter 300 receiving the ultra low frequency band information generates a target frequency signal Tf for the ultra low frequency band VLF. The heart rate variance analyzer 400 generates a heart rate variance signal HRV for the ultra low frequency band VLF based on the target frequency signal Tf.

The target frequency band mode may change over time. For example, the band selector 500 operates in the high frequency band mode for a predetermined time to provide target frequency band information fti for the high frequency band HF, and thereafter, by changing the target frequency band mode, the low frequency band. Target frequency band information fti for LF may be provided. Accordingly, by changing the mode of the band selector 500, the frequency converter 300 may generate target frequency signals Tf for various frequency bands. In addition, based on this, the heart rate variability analysis unit 400 generates various heart rate variability signals (HRV) to enable multi-faceted analysis of heart rate variability.

The storage unit 600 may be used as a storage device for the heart rate variability analyzer 400. The storage unit 600 receives and stores the heart rate variance signal HRV, the target frequency signal Tf, and the target frequency band information fti from the heart rate variability analyzer 400.

The storage unit 600 may include a read only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable and programmable ROM (EEPROM), a flash memory device, a phase-change RAM (PRAM), and a magnetic magnetic memory (MRAM). It may include at least one of a nonvolatile memory device such as RAM (RRAM), Resistive RAM (RRAM), or Ferroelectric RAM (FRAM), and may include a static RAM (SRAM), a dynamic RAM (DRAM), or a synchronous DRAM (SDRAM). It may include at least one of the volatile memory device.

The display unit 700 may include a heart rate variance signal HRV generated by the heart rate variability analyzer 400, a target frequency signal Tf generated by the frequency converter 300, or data stored in the storage 600. Based on the received image signal (IHRV) can be displayed. The display unit 700 may include at least one of a liquid crystal display (LCD), an organic light emitting diode (OLED), an active matrix OLED (AMOLED), a flexible display, or an electronic ink.

4 is a block diagram of the frequency converter 300 according to an embodiment of the present invention. 5 to 8 are graphs for explaining a process of detecting the signal of the target frequency band by the frequency converter 300 of FIG.

Referring to FIG. 4, the frequency converter 300 includes frequency shifters 311 and 312, low pass filters 321 and 322, decimators 331 and 332, data inserters 341 and 342, and high speed. Fourier transform unit 350 is included. However, unlike FIG. 4, the frequency converter 300 may not include some of the components. For example, the frequency converter 300 may not include the decimators 331 and 332 or may not include the data inserters 341 and 342.

The frequency shifters 311 and 312 include a first frequency shifter 311 and a second frequency shifter 312. The first frequency shifter 311 and the second frequency shifter 312 receive the RR interval time series signal RPn from the R-peak detector 200.

The first frequency shifter 311 receives the RR interval time series signal RPn to generate an in-phase (I) channel signal. The first frequency shifter 311 multiplies the first sine signal Ift by the RR interval time series signal RPn to generate an I-channel signal. When the RR interval time series signal RPn is a discrete signal for n, the first sinusoidal signal Ift is defined as cos (2 * pi * fc * n * ts). The fc of the first sinusoidal signal Ift is defined as the center frequency of the target frequency band ft to be analyzed. Ts of the first sinusoidal signal Ift corresponds to the inverse of the sampling frequency.

The second frequency shifter 312 receives the RR interval time series signal RPn to generate a quadrature-phase (Q) channel signal. The second frequency shifter 312 multiplies the second sine signal Qft by the RR interval time series signal RPn to generate a Q channel signal. When the RR interval time series signal RPn is a discrete signal for n, the second sinusoidal signal Qft is defined as -sin (2 * pi * fc * n * ts). The first frequency shifter 311 and the second frequency shifter 312 classify the RR interval time series signal RPn into an I-channel signal and a Q-channel signal. The first frequency shifter 311 extracts the real part of the RR interval time series signal RPn, and the second frequency shifter 312 extracts the imaginary part of the RR interval time series signal RPn.

The frequency converter 300 may further include a classifier (not shown) for generating the first sinusoidal signal Ift and the second sinusoidal signal Qft to classify the RR interval time series signal RPn. The classifier may generate a first sinusoidal signal Ift and a second sinusoidal signal Qft. Since the frequencies of the first sinusoid signal Ift and the second sinusoidal signal Qft are based on the target frequency band ft and the sampling frequency, the classifier can receive the target frequency band information fti from the band selector 500. Can receive the sampling signal fs. Alternatively, the classifier may be provided to the band selector 500 to provide the first sinusoidal signal Ift and the second sinusoidal signal Qft to the first frequency shifter 311 and the second frequency shifter 312. Can be. Alternatively, the classifier may be provided as a separate component in the heart rate variability analysis apparatus 1000.

When the first sinusoidal signal Ift or the second sinusoidal signal Qft is multiplied by the RR interval time series signal RPn, the RR interval time series signal RPn moves in the frequency domain by the center frequency fc. That is, the target frequency band ft moves to baseband. The first frequency shifter 311 and the second frequency shifter 312 classify the RR interval time series signal RPn into an I-channel signal and a Q-channel signal, and shift the target frequency band ft to the baseband. The first sinusoid signal Ift and the second sinusoid signal Qft are respectively multiplied, and a shift detection signal is generated.

The first low pass filter 321 and the second low pass filter 322 receive the shift detection signal. The first low pass filter 321 receives the I channel signal, and the second low pass filter 322 receives the Q channel signal. The first low pass filter 321 and the second low pass filter 322 pass the baseband shift detection signal and block the high frequency band shift detection signal. Therefore, a signal corresponding to the target frequency band ft may be selectively extracted from the RR interval time series signal RPn. In addition, since the signal of the frequency band to be analyzed and the signal of the remaining bands are cut off, unnecessary data can be removed to reduce the amount of data.

The first low pass filter 321 and the second low pass filter 322 may convert the cutoff frequency according to the target frequency band mode of the band selector 500. For example, the high frequency bandwidth is about 0.25 Hz in the high frequency band mode, the low frequency bandwidth is about 0.11 Hz in the low frequency band mode, and the ultra low frequency bandwidth is about 0.04 Hz in the ultra low frequency band mode. Therefore, a cutoff frequency of about 0.125 Hz is required in the high frequency band mode, a cutoff frequency of about 0.055 Hz is required in the low frequency band mode, and a cutoff frequency of about 0.02 Hz may be required in the ultra low frequency band mode. Accordingly, the first low pass filter 321 and the second low pass filter 322 may receive the target frequency band information fti from the band selector 500 to convert the cutoff frequency. The target frequency band information fti may include bandwidth information fl for a corresponding target frequency band mode.

The first decimator 331 receives the shift detection signal passing through the first low pass filter 321, and the second decimator 332 passes the shift detection signal passing through the second low pass filter 322. Receive The first decimator 331 and the second decimator 332 reduce the sampling rate of the signal. For example, when the first decimator 331 and the second decimator 332 perform the fourth decimation, the sampling frequency is reduced by 1/4 of the existing sampling frequency. In this case, the data of the shift detection signal can be reduced by 1/4. In addition, the baseband shift detection signal may appear as a waveform extended four times based on the sampling frequency in the frequency domain. When the first decimator 331 and the second decimator 332 perform M-order decimation, the frequency resolution is determined based on the M value. In performing M-order decimation, the frequency resolution is defined as (sampling frequency / M) / (number of FFT points).

The first decimator 331 and the second decimator 332 may receive the magnification signal MM to determine the order of decimation. Although not shown, the frequency converter 300 may further include a multiplier that generates a magnification signal MM and provides a magnification signal MM to the first decimator 331 and the second decimator 332. . Alternatively, the multiplier may be provided as a separate component in the heart rate variability analyzer 1000.

The first data inserter 341 and the second data inserter 342 zero-pad the zero data to the decimated signal. Data of zero may be inserted between the data of each of the shift detection signals. The first data inserting unit 341 and the second data inserting unit 342 have a reference FFT when the number of data output from the first and second decimators 331 and 332 is directly input to the fast Fourier transform unit 350. When the number of points is smaller than the number of points, by inserting zero data by the difference in the number of data, the frequency resolution can be improved and the precision of the signal can be improved.

The first data inserter 341 and the second data inserter 342 may receive the padding signal PP to determine the number of data insertions. Although not shown, the frequency converter 300 further includes a resolution adjuster that generates a padding signal PP and provides the padding signal PP to the first data inserter 341 and the second data inserter 342. can do. Alternatively, the resolution adjuster may be provided as a separate component in the heart rate variability analyzer 1000.

The first decimator 331 and the second decimator 332 can concentrate and observe the RR interval time series signal RPn for the target frequency band ft and reduce the amount of data. The first data inserter 341 and the second data inserter 342 insert data to improve frequency resolution of the RR interval time series signal RPn for the target frequency band ft and to improve accuracy. Accordingly, the first data inserter 341 and the second data inserter 342 may be eliminated in view of data efficiency and processing speed.

The fast Fourier transform unit 350 receives a signal into which zero data is inserted from the first data inserter 341 and the second data inserter 342. The fast Fourier transform unit 350 uses the first and second low pass filters 321 and 322, the first and second decimators 331 and 332, and the first and second data inserters 341 and 342. Fast Fourier Transform the transmitted RR interval time series signal (RPn). The fast Fourier transform unit 350 includes a fast Fourier transform algorithm for reducing the complicated calculation of the Discrete Fourier Transform, and generates the target frequency signal Tf based on the fast Fourier transform algorithm.

5 exemplarily illustrates a signal input to the frequency converter 300. 5 is a graph showing frequency conversion of a signal input to the frequency converter 300. The user may input the target frequency band information fti to the frequency converter 300 to observe the target frequency band ft.

FIG. 6 illustrates the magnitude of a signal in which the signal of FIG. 5 is converted by the first frequency shifter 511 and the second frequency shifter 512. The first frequency shifter 511 or the second frequency shifter 512 multiplies the signal of FIG. 5 by the first sinusoidal signal Ift or the second sinusoidal signal Qft to base the signal of the target frequency band ft. Move the band. In the frequency domain, the signal of FIG. 5A moves in the negative direction by the center frequency fc.

FIG. 7 illustrates the magnitudes of the signals of FIG. 6 converted by the first and second low pass filters 321 and 322. The first and second low pass filters 321 and 322 may have a cutoff frequency value of a target frequency band ft / 2. By the first and second low pass filters 321 and 322, only the baseband waveform remains, and the remaining waveforms are removed.

FIG. 8 illustrates the magnitudes of the signals of FIG. 7 converted by the first and second decimators 331 and 332. The first and second decimators 331 and 332 perform M order decimation. By decimation, the sampling frequency is reduced by 1 / M. Accordingly, the first and second decimators 331 and 332 allow the signals of the target frequency band ft to be concentrated as in the signal of FIG. 8.

9 is a block diagram of an apparatus for analyzing heart rate variability according to another embodiment of the present invention.

Referring to FIG. 9, the heart rate variability analyzer 2000 includes a converter 2100, an R-peak detector 2200, a first frequency converter 2300a, a second frequency converter 2300b, and a third frequency converter. 2300c, and heart rate variability analyzer 2400. The configuration of the converter 2100, the R-peak detector 2200, and the heart rate variability analyzer 2400 of FIG. 6 is respectively represented by the converter 100, the R-peak detector 200, and the heart rate variability analyzer of FIG. 3. 400, detailed description thereof will be omitted.

The first frequency converter 2300a receives the RR interval time series signal RPn from the R-peak detector 2200 and outputs a first target frequency signal Tf1. The first frequency converter 2300a extracts a signal for the first target frequency band ft1 among the RR interval time series signals RPn. For example, the first target frequency band ft1 may be a high frequency band HF.

The second frequency converter 2300b receives the RR interval time series signal RPn from the R-peak detector 2200 and outputs a second target frequency signal Tf2. The second frequency converter 2300b extracts a signal for the second target frequency band ft2 among the RR interval time series signals RPn. For example, the second target frequency band ft2 may be a low frequency band LF.

The third frequency converter 2300c receives the RR interval time series signal RPn from the R-peak detector 2200 and outputs a third target frequency signal Tf3. The third frequency converter 2300c extracts a signal for the third target frequency band ft3 from the RR interval time series signal RPn. For example, the third target frequency band ft3 may be an ultra low frequency band VLF.

Compared to the heart rate variability analysis apparatus 1000 of FIG. 3, the heart rate variability analysis apparatus 2000 of FIG. 9 includes a plurality of frequency converters. Therefore, the heart rate variability signals of various frequency bands can be analyzed simultaneously. For example, the heart rate variability analyzer 2000 may simultaneously calculate and analyze heart rate variability of the high frequency band HF, the low frequency band LF, and the ultra low frequency band VLF. Compared to the heart rate variability analysis apparatus 2000 of FIG. 9, the heart rate variability analysis apparatus 1000 of FIG. 3 may analyze the heart rate variability with a relatively small data capacity.

10 is a flowchart illustrating a method for detecting heart rate variability (S1000) according to an embodiment of the present invention.

Referring to FIG. 10, the heart rate variance detection method S1000 may include receiving an electrocardiogram signal (S100), detecting an R-peak value (S200), converting a frequency (S300), and receiving a heart rate variance signal. Computing and analyzing step (S400).

In operation S100 of receiving an ECG signal, the heart rate variability detection apparatus receives an analog ECG signal ECGT from the outside. Receiving the ECG signal (S100) may include converting an analog ECG signal ECGT into a digital ECG signal ECGn. Receiving the ECG signal (S100) is performed by the analog-to-digital converter (100, 2100). In this case, the analog ECG signal ECGT may be converted into the digital ECG signal ECGn based on the sampling frequency included in the sampling signal fs received by the analog-digital converters 100 and 2100.

In operation S200, the heart rate variance detection device detects the R-peak value of the ECG signal. The step S200 of detecting the R-peak value is performed by the R-peak detectors 200 and 2200. The detecting of the R-peak value (S200) may include generating an RR interval time series signal RPn based on the R-peak value. The RR interval time series signal RPn may be generated by calculating the RR interval RRI based on sequentially detected R-peak values.

In operation S300, the heart rate variability detection apparatus extracts a target frequency band ft of the RR interval time series signal RPn and converts the target frequency band ft into a target frequency signal Tf. The step of converting the frequency (S300) is performed by the frequency converter 300 of FIG. 3 or the first to third frequency converters 2300a, 2300b, and 2300c of FIG. 9. Specific details will be described later.

In operation S400 of calculating and analyzing the heart rate variance signal, the heart rate variance detection apparatus generates a heart rate variance signal for the target frequency band ft based on the target frequency signal Tf. Computing and analyzing the heart rate variance signal (S400) is performed by the heart rate variability analyzer 400 and 2400. In the step of calculating and analyzing the heart rate variance signal (S400), the heart rate variability analysis units 400 and 2400 may output power of the high frequency band HF, power of the low frequency band LF, or power of the ultra low frequency band VLF. Values, power ratios (LF / HF), etc. can be calculated and analyzed to determine the activity of the autonomic nervous system and the like, and mental health can be progressed.

FIG. 11 is a flowchart illustrating a step S300 of converting a frequency of FIG. 10.

Referring to FIG. 11, the step of converting the frequency (S300) includes classifying the Q channel and the I channel (S310), moving to the baseband (S320), filtering with a low pass filter (S330), and desiccation. Imaging (S340), zero padding (S350), and fast Fourier transforming (S360). Each step of FIG. 8 is performed by the frequency converter 300 of FIG. 3 or the first to third frequency converters 2300a, 2300b, and 2300c of FIG. 9.

In the step S310 of classifying the Q channel and the I channel, the apparatus for analyzing heart rate variance classifies the RR interval time series signal RPn into an I channel corresponding to the real part and a Q channel corresponding to the imaginary part. The step S310 of classifying the Q channel and the I channel may include generating an I channel by multiplying the first sinusoid signal Ift by the RR interval time series signal RPn and the second sinusoidal signal Qft by the RR interval time series signal ( Multiplying RPn) to generate a Q channel. The generating of the I channel is performed by the first frequency moving unit 311, and the generating of the Q channel is performed by the second frequency moving unit 312.

In operation S320, the heart rate variability analysis apparatus moves the target frequency band ft of the RR interval time series signal RPn to the base band to generate a shift detection signal. The step S320 of moving to the base band is performed by the first frequency moving unit 311 and the second frequency moving unit 312. In step S320, the RR interval time series signal RPn moves by the center frequency fc of the target frequency band ft. The first sinusoidal signal Ift and the second sinusoidal signal Qft may be a sinusoidal function to move to the baseband, and the frequency of the sinusoidal function is proportional to the center frequency fc. The step of classifying the Q channel and the I channel (S310) and the step of moving to the base band (S320) may be performed simultaneously in the first frequency moving unit 311 and the second frequency moving unit 312.

In the step S330 of filtering with the low pass filter, the heart rate variability analyzer passes the baseband shift detection signal and removes the shift detection signal of the remaining frequency band. The step S330 of filtering with the low pass filter is performed by the first and second low pass filters 321 and 322.

In operation 340, the heart rate variability analyzer decreases the sampling rate of the filtered shift detection signal. The decimation step S340 is performed in the first and second decimators 331 and 332. In the decimation step S340, the amount of data is reduced, and the sampling frequency is reduced by the decimation order.

In the zero padding step S350, the heart rate variability analyzer inserts zero data into the filtered shift detection signal. The zero padding step S350 is performed by the first and second data insertion units 341 and 342. When the number of data input to the fast Fourier transform unit 350 is smaller than the number of reference FFT points when the data is decreased by the step 340 of decimating in step S350, the first and second data are insufficient. The inserting units 341 and 342 insert zero data by the difference in the number of data. In the zero padding step S350, the amount of data is increased and the frequency resolution is improved.

In a fast Fourier transform step (S360), the heart rate variability analyzer performs a fast Fourier transform of the filtered, decimated, and zero-padded shifting detection signal from the time domain to the frequency domain. The fast Fourier transforming step S360 is performed by the fast Fourier transforming unit 350.

The above description is specific examples for practicing the present invention. The present invention will include not only the embodiments described above but also embodiments that can be easily changed or simply changed in design. In addition, the present invention will also include techniques that can be easily modified and carried out using the above-described embodiments.

1000, 2000: heart rate variability analyzer 100, 2100: converter
200, 2200: R-peak detector 300, 2300a, 2300b, 2300c: frequency converter
400, 2400: heart rate variability analysis unit 500: band selector
fs: sampling signal ft: target frequency band
fc: center frequency ECGT: analog ECG signal
ECGn: Digital ECG Signal RPn: RR Interval Time Series Signal
Tf: Target Frequency Signal HRV: Heart Rate Variability Signal

Claims (16)

A converter for receiving an analog electrocardiogram signal comprising a QRS wave and converting it into a digital electrocardiogram signal;
An R-peak detector configured to receive the digital electrocardiogram signal, detect an R-peak value, and generate an RR interval time series signal based on the R-peak value;
A frequency converter configured to extract a signal of a target frequency band from the RR interval time series signal to generate a target frequency signal; And
A heart rate variance analyzer configured to generate a heart rate variance signal for the target frequency band based on the target frequency signal;
The frequency converter,
A frequency shifter configured to shift the RR interval time series signal by a reference frequency within the target frequency band to generate a shift detection signal;
A low pass filter for passing the shift detection signal at a cutoff frequency dependent on a bandwidth of the target frequency band; And
And a fast Fourier transform unit generating a target frequency signal based on the shift detection signal passing through the low pass filter. The method of claim 1,
The frequency moving unit,
A heart rate variability analysis device for generating the shift detection signal comprising an in-phase (I) channel signal and a quadrature-phase (Q) channel signal. The method of claim 2,
The frequency moving unit,
Generating an I channel signal by multiplying the RR interval time series signal by a first sinusoidal signal, generating a Q channel signal by multiplying the RR interval time series signal by a second sinusoidal signal,
And the first sinusoidal signal and the second sinusoidal signal have a phase difference of 90 degrees, and a frequency of the first sinusoidal signal and the second sinusoidal signal is based on a center frequency of the target frequency band. The method of claim 1,
The converter,
Heart rate variability analysis device for generating the digital electrocardiogram signal sampled at a sampling frequency. The method of claim 4, wherein
The frequency converter,
And a decimator for reducing a sampling rate of the shift detection signal passing through the low pass filter. The method of claim 4, wherein
The frequency converter,
And a data insertion unit for inserting zero data into the shift detection signal passing through the low pass filter. The method of claim 1,
And a band selector configured to provide target frequency band information corresponding to a target frequency band mode in the frequency converter. The method of claim 7, wherein
The target frequency band mode is,
Including a high frequency band mode, a low frequency band mode, and an ultra low frequency band mode,
The band selector,
Providing the high frequency band information to the frequency converter in the high frequency band mode, providing the low frequency band information to the frequency converter in the low frequency band mode, and providing the ultra low frequency band information to the frequency in the ultra low frequency band mode Heart rate variability analysis device provided to the conversion unit. The method of claim 8,
The target frequency band includes a high frequency band,
The high frequency band is heart rate variability analysis apparatus comprising a band between 0.15 to 0.4 Hz. The method of claim 8,
The target frequency band includes a low frequency band,
The low frequency band is heart rate variability analysis apparatus comprising a band between 0.04 to 0.15 Hz. The method of claim 8,
The target frequency band includes an ultra low frequency band,
The ultra low frequency band heart rate variability analysis apparatus comprising a band between 0.003 to 0.04 Hz. The method of claim 1,
The heart rate variability signal is a heart rate variability analysis apparatus including a power spectrum signal of the target frequency band. The method of claim 12,
The target frequency band includes a high frequency band and a low frequency band,
The heart rate variability analysis unit calculates at least one of the power value of the high frequency band, the power value of the low frequency band, and the power ratio of the low frequency band to the high frequency band based on the power spectrum signal. Receiving an electrocardiogram signal;
Detecting an R-peak value of the ECG signal;
A frequency conversion step of generating a target frequency signal by extracting a target frequency band of the RR interval time series signal generated based on the R-peak value; And
Calculating and analyzing a heart rate variance signal for the target frequency band based on the target frequency signal,
The frequency conversion step,
Shifting the RR interval time series signal by a reference frequency within the target frequency band;
Filtering the RR interval time series signal shifted by the reference frequency with a low pass filter having a cutoff frequency dependent on the bandwidth of the target frequency band; And
Fast Fourier transforming the filtered RR interval time series signal. The method of claim 14,
The frequency conversion step,
And classifying the RR interval time series signal into an in-phase (I) channel signal and a quadrature-phase (Q) channel signal. The method of claim 15,
The frequency conversion step,
Reducing a sampling rate of the filtered RR interval time series signal; And
And inserting zero data into the filtered RR interval time series signal.

Images