CN1271561A - Data compression method and device for Holter system - Google Patents

Abstract

A data compression method and its device for Holter system are disclosed. An integer-integer small wave transform method and its multi-resolution analysis technique are used to decompose the cardioelectric signals by multi-scale small waves. The start position of compound QRS wave is determined by a compound QRS wave detector circuit, so that the shield function can be calculated. The decomposing classes and error threshold are chosen in advance according to data sampling rate and required compression effect. The small wave components at different scales after small wave is decomposed are compared with the product of error threshold and shield function point by point to determine the preserved points, and then the recorded preserved data are further transmitted to a computer.

Description

The data compression method of Holter system and device thereof

The invention belongs to the Medical Instruments technical field, further relate to the method and the relevant apparatus of compression storage electrocardiosignal waveform in the Holter system.

The recorder that was used to write down body surface ecg in 24 hours in the digital continuous record formula Holter system is carried by patient, because the recorder volume is as far as possible little, the disposal ability of institute's employing processor (CPU) and the capacity of memorizer all are subjected to strict restriction, generally will carry out compression storing data to the electrocardiosignal that is recorded by body surface.Since the performance of general processor (CPU) a little less than, be the real-time that guarantees data compression, compression algorithm is generally the direct compression method of the time domain that can realize fast.The direct compression method of time domain is to be based upon on the superfluous abundant basis of direct analysis initial data, and data point is divided into retention point and superfluous abundant point, by rejecting the purpose that superfluous abundant point reaches data compression.Generally use broken line reconstruct when data reconstruction, the waveform that obtains is stiff, has had a strong impact on its clinical value; Be difficult to guarantee high data compression rate and high information fidelity degree simultaneously do not have the such high frequency of the reservation ventricular late potential ability of information by a narrow margin in addition, and very responsive to High-frequency Interference.

The objective of the invention is to overcome the shortcoming of above-mentioned prior art, it is whole to propose a kind of employing---and whole wavelet transformation is realized the method that efficient data compresses to electrocardiosignal, and has more high performance Holter product with this method exploitation.

Fig. 1 is a structured flowchart of the present invention.

Fig. 2 is electrocardiosignal amplifying circuit circuit theory diagrams of the present invention.

Fig. 3 is a QRS complex wave testing circuit schematic diagram of the present invention.

Below in conjunction with accompanying drawing principle of the present invention is elaborated.

Referring to Fig. 1, the inventive system comprises an electrocardiosignal amplifying circuit (1), a QRS complex wave testing circuit (2), an A/D change-over circuit (3), a CPU (4) and a memory circuit (5).The body surface signal input ecg signal amplifying circuit (1) that measures by body surface, A/D change-over circuit (3) is sent in the output of electrocardiosignal amplifying circuit (1), the output of electrocardiosignal amplifying circuit (1) is simultaneously sent into QRS complex wave testing circuit (2) as its input, the interrupt signal of CPU (4) as CPU (4) sent in the output of QRS complex wave testing circuit (2), the output of A/D change-over circuit (3) links to each other with the bus of CPU (4), and the bus of CPU (4) also links to each other with memory circuit (5) simultaneously.

Referring to Fig. 2, electrocardiosignal amplifying circuit of the present invention (1) comprises the second order high-pass filtering amplifier of being made up of A1, R1, R2, R3, R4, C1, C2, the body surface signal is imported this second order high-pass filtering amplifier, the second-order low-pass filter amplifier of being made up of A2, R5, R6, R7, R8, C3, C4 is sent in its output, and A/D change-over circuit (3) is sent in the output of this second-order low-pass filter amplifier.

Referring to Fig. 3, QRS complex wave testing circuit of the present invention (2) comprises the band filter of being made up of A3, R9, R10, R11, C5, C6, its output is received the anode of diode D2 respectively and the input of the phase inverter be made up of A4, R15, R16, the anode of diode D1 is received in the output of this phase inverter, the negative electrode of D1 links to each other with the negative electrode of D2 and receives the IN+ end of IC1 comparator, IC1, R12, R13, R14 form QRS complex wave detection comparator, the IC1 comparator the OUT end interrupt importing with one of CPU (4) and link to each other.

The method that the present invention is used for data compression is: the frequency band of electrocardiosignal amplifying circuit (1) is 0.03---200Hz, the body surface signal obtains electrocardiosignal after by electrocardiosignal amplifying circuit (1), and QRS complex wave testing circuit (2) detects the QRS complex wave and sends interrupt signal in real time to CPU (4).

The ecg signal data that sampling obtains is with corresponding the putting in order of formula (A)---whole wavelet transformation decomposes. $b_{2^m}^{\′}\left(k\right)=a_{2^{m+1}}(2k+1)-a_{2^{m+1}}\left(2k\right)$ $a_{2^m}\left(k\right)=a_{2^{m+1}}\left(2k\right)+[b_{2^m}^{\′}\left(k\right)/2],----\left(A\right)$ $b_{2^m}\left(k\right)=b_{2^m}^{\′}\left(k\right)+[a_{2^m}(k-1)/4-a_{2^m}(k+1)/4+1/2]$

Wherein Be input signal, With

Be respectively low frequency and the high frequency output signal that decomposition obtains, Be the intermediate variable in the computational process,

Representative is no more than the maximum integer of x.Can be by this mapping original input signal

Be decomposed into the high-frequency signal that length is respectively half And low frequency signal

Adopt the multiresolution analysis technology, to the low frequency signal after decomposing With same process at different yardstick 2 ^M+1Decompose (M≤m≤-1, M is the maximum progression that desire is decomposed), can be decomposed into one group of Wavelet Component on different wavelet scales to primary signal.

Each heart beat cycle of electrocardiosignal mainly comprises useful informations such as QRS complex wave, P ripple, T ripple, ST section and ventricular late potential, and they appear at the different moment of the particular dimensions after the above-mentioned wavelet decomposition respectively.Clinical practice information for both can complete reservation electrocardiosignal can obtain high as far as possible data compression rate again.In compression process, must carry out different processing to transform data at diverse location, different scale.Introduce mask function, it gets different value in the different sections of signal: in QRS complex wave part value is 2, partly is 1 in ventricular late potential, and other parts are 3 for this reason.By QRS complex wave testing circuit (2) detect the QRS complex wave and to CPU (4) thus sending interrupt signal in real time determines QRS complex wave original position.

Compression effectiveness according to data sampling rate and needs is selected decomposed class and error threshold in advance, to comparing through the Wavelet Component pointwise on different wavelet scales after the above-mentioned wavelet decomposition and the product of error threshold and mask function, if keep this point greater than this product, otherwise abandon.

Identical with above-mentioned catabolic process, with formula (C) restructural compressed waveform. $b_{2^m}^{\′}\left(k\right)=b_{2^m}\left(k\right)-[a_{2^m}(k-1)/4-a_{2^m}(k+1)/4+1/2]$ $a_{2^{m+1}}\left(2k\right)=a_{2^m}\left(k\right)-[b_{2^m}^{\′}\left(k\right)/2]---\left(C\right)$ $a_{2^{m+1}}(2k+1)=a_{2^{m+1}}\left(2k\right)+b_{2^m}^{\′}\left(k\right)$

Embodiments of the invention are: the Holter system of exploitation 1000Hz electrocardiosignal sample rate.By body surface detection to the body surface signal obtain electrocardiosignal after by electrocardiosignal amplifying circuit (1), with this electrocardiosignal of 1000Hz sample rate real-time sampling, QRS complex wave testing circuit (2) detects the QRS complex wave and sends interrupt signal in real time to CPU (4), thereby calculates mask function so that determine QRS complex wave starting point.The electrocardiosignal that sampling obtains is carried out segmentation by every section 2048 sampled points, with above-mentioned data compression method electrocardiosignal is carried out the Real Time Compression code storage to every section then, thereby finish 24 hours record of electrocardiosignal.After record is finished the transfer of data that records is arrived computer, thereby realize 24 hours arrhythmia analysis, ST piecewise analysis, QT analysis, heart rate variability analysis and ventricular late potential dynamic analysis.

The data compression method of the Holter system that the present invention proposes not only can complete reservation QRS, information such as P, T, ST, and can keep ventricular late potential information effectively.This method is owing to adopt whole---and whole wavelet transformation has been avoided the floating-point operation problem, only needs to add, subtracts and shift operation, can realize fast, is particularly suitable for developing the high-performance Holter system with ventricular late potential record and analysis ability.The thought that proposes with the present invention adopts other whole---and whole wavelet transformation extends to the compression storage of other biomedicine signals even the data compression of other subjects.

Claims (5)

1.Holter the data compression method of system, the body surface signal obtains electrocardiosignal after by electrocardiosignal amplifying circuit (1), characteristics of the present invention are, detect the QRS complex wave and send interrupt signal in real time with QRS complex wave testing circuit (2) to CPU (4), compression effectiveness according to data sampling rate and needs is selected decomposed class and error threshold in advance, to comparing through the Wavelet Component pointwise on different wavelet scales after the above-mentioned wavelet decomposition and the product of error threshold and mask function, if keep this point greater than this product, otherwise abandon, and then the transfer of data that records is arrived computer.

2. method according to claim 1, it is characterized in that, by body surface detection to the body surface signal obtain electrocardiosignal after by electrocardiosignal amplifying circuit (1) after, with this electrocardiosignal of 1000Hz sample rate real-time sampling, QRS complex wave testing circuit (2) detects the QRS complex wave and sends interrupt signal in real time to CPU (4), thereby calculate mask function so that determine QRS complex wave starting point, the electrocardiosignal that sampling obtains is carried out segmentation by every section 2048 sampled points, with above-mentioned data compression method electrocardiosignal is carried out the Real Time Compression code storage to every section then, thereby finish 24 hours record of electrocardiosignal, after record is finished the transfer of data that records is arrived computer, thereby realize 24 hours arrhythmia analysis, the ST piecewise analysis, QT analyzes, heart rate variability analysis, with the ventricular late potential dynamic analysis.

3. implement the device of the described method of claim 1, comprise an electrocardiosignal amplifying circuit (1), a QRS complex wave testing circuit (2), an A/D change-over circuit (3), a CPU (4), it is characterized in that the body surface signal input ecg signal amplifying circuit (1) that measures by body surface with a memory circuit (5), A/D change-over circuit (3) is sent in the output of electrocardiosignal amplifying circuit (1), the output of electrocardiosignal amplifying circuit (1) is simultaneously sent into QRS complex wave testing circuit (2) as its input, the interrupt signal of CPU (4) as CPU (4) sent in the output of QRS complex wave testing circuit (2), the output of A/D change-over circuit (3) links to each other with the bus of CPU (4), and the bus of CPU (4) also links to each other with memory circuit (5) simultaneously.

4. device according to claim 3, it is characterized in that, said electrocardiosignal amplifying circuit (1) comprises the second order high-pass filtering amplifier of being made up of A1, R1, R2, R3, R4, C1, C2, the body surface signal is imported this second order high-pass filtering amplifier, the second-order low-pass filter amplifier of being made up of A2, R5, R6, R7, R8, C3, C4 is sent in its output, and A/D change-over circuit (3) is sent in the output of this second-order low-pass filter amplifier.

5. device according to claim 3, it is characterized in that, said QRS complex wave testing circuit (2) comprises the band filter of being made up of A3, R9, R10, R11, C5, C6, its output is received the anode of diode D2 respectively and the input of the phase inverter be made up of A4, R15, R16, the anode of diode D1 is received in the output of this phase inverter, the negative electrode of D1 links to each other with the negative electrode of D2 and receives the IN+ end of IC1 comparator, IC1, R12, R13, R14 form QRS complex wave detection comparator, the IC1 comparator the OUT end interrupt importing with one of CPU (4) and link to each other.

Images