CN1404794A - Wide-band transmission dynamic electrocardiograph - Google Patents

Abstract

The broad-band HOLTER system comprises HOLTER recorder, radio-frequency modem, radio-frequency emitter, radio frequency receiver, radio-frequency modem, PC machine, ADSL modem, telephone network, telephone exchange, broad-band network, ADSL modem and PC computer.

Description

The dynamic ecg of wideband transmit

The present invention relates to a kind of broadband ambulatory ecg signal record, transmission, the method for analysis.

Ambulatory electrocardiogram was that U.S. Norman J doctor Holter initiated in nineteen fifty-seven, dropped into clinical use in 1961, was called for short HOLTER.HOLTER can once write down 24 hours, the electrocardiogram of patient in normal activity.And carry out arrhythmia, the analysis of ST section (myocardial ischemia) and noninvasive prognosis by computer.Be in the hospital, detect and the diagnosis of cardiovascular diseases necessary means.

Traditional HOLTER is mainly by electrocardiorecorder and computer, and analysis software package constitutes.

1. the arrhythmic events analysis that takes place in 24 hours, to judge ARR frequency (24 hours or ARR number of times per hour takes place), ARR character, whether taken place serious fatefulue arrhythmia (as ventricular tachycardia, paired ventricular premature contraction, R-ON-T, polynary source property ventricular premature contraction, Dou Xingting is rich, bradycardia, have or not the sinuatrial node pathological changes, whether the indication of installation pacemaker etc. is arranged).

2. the variation and the morphological analysis of the ST section in 24 hours is to judge myocardial ischemia, especially asymptomatic myocardial ischemia.With patient's daily record combination, can judge the motivating factor that causes myocardial ischemia.

3. owing to be the non-volatile recording electrocardiogram, can reflect patient's (different physical load, under different emotions and the autonomic nerve state, drug effect is inferior) under different conditions, the regularity that arrhythmia and myocardial ischemia take place and other relation extraneous and oneself factor.Thereby carry out corresponding treatment.

In a word, traditional HOLTER is using under regular hour, the space: promptly under doctor's operation, wear and take off recording box to patient in hospital.Can only use 24 hours.The present invention plans to build upright one and is enclosed within infinite space, operable HOLTER under the time.

The present invention is based on broadband networks and has entered huge numbers of families.

Since the present invention was based on 99 years, the large-scale circuit memorizer (FLASH MEMORY) of jumbo (monolithic 8 Mbytes-16 Mbytes) occurs, reach 256MBIT (16MBYTE) monolithic, can realize undistorted record, (FLASH MEMORY) is the Holter of recording medium with the large-scale circuit memorizer, can write down and analyze 12 electrocardiograms that lead; Adopt data compression algorithm in addition, further improve the memory space of Holter, reduce the volume of recorder, improve sample rate, reduce cost.

Automatic identification abnormal electrocardiogram and its statistics aspect that the present invention is based on Holter are ripe.

The present invention is based on the diagnostic value that medical circle has been generally acknowledged Holter.

Practice for many years confirms that electrocardiogram is the most effective means of diagnosis of cardiovascular diseases.But general electrocardiograph is just measured patient under the state of tranquillization, the electrocardiogram in several seconds.Be difficult to pounce on and grasp an Electrocardiographic anomalous event of crossing property (the QT interval of short battle array prolongs other asymptomatic electrocardiographic abnormality etc. for the electrocardiographic abnormality during as angina pectoris, short battle array ventricular tachycardia).The phone transmission electrocardiogram has had very big progress than general electrocardiograph, but still has its limitation.Because the Electrocardiographic record of phone transmission is not in good time by the own subjectivity of patient, by patient oneself operation, be difficult to write down patient's subjective sensation less than electrocardiographic abnormality (as silent ischemia).Also be difficult to ARR attack degree (as the number of times of ventricular premature contraction per hour) and frequency are carried out quantitative estimation.In numerous arrhythmic events, also be difficult to classify for arrhythmic events, pounce on and grasp grade of malignancy than higher arrhythmic events (as paired ventricular premature contraction, polyphyly ventricular premature contraction, R-ON-T, long QT interval etc.).

The limitation of general electrocardiograph (the phone transmission electrocardiogram that comprises short record):

1. be difficult to pounce on Electrocardiographic unusual (the descending QT interval prolongation etc. as asymptomatic ST section) that catch the Electrocardiographic unusual of a property crossed, particularly patient's subjective symptom unconspicuous one property crossed.

2. be difficult to the ARR frequency of more quantitative estimation and the order of severity (as how many ventricular premature contractions per hour being arranged, how many times sinus arrest etc.).

3. be difficult to find and the ARR character of classifying fatefulue to find, serious arrhythmia.As the polyphyly ventricular premature contraction, paired ventricular premature contraction, R-ON-T, paroxysmal atrial fibrillation, long QT interval, the fast mexiletine that the sinuatrial node pathological changes causes etc.

4. be difficult to find silent ischemia.

5. be difficult to more quantitative Electrocardiographic unusual and its motivating factor be connected.Excite the factor of arrhythmia and myocardial ischemia generally to have: the multiple lotus of muscle power is increased the weight of, direct stimulation, autonomic unbalance (sympathetic activation), heavy meal is taken medicine etc.When above-mentioned motivating factor occurred, HOLTER can be relatively easy to diagnose out fatefulue, serious arrhythmic events whether to increase (what have increased), and whether ST section ischemia is more serious or the like.Particularly patient's subjective sensation less than abnormal electrocardiographic pattern (as phenomenons such as asymptomatic myocardial ischemia).

6.HOLTER can be used for comparison quantitative assessment medicine effect.As what (per hour) the back arrhythmia anomalous event of taking medicine reduced, and whether the ST section is improved etc.

7.HOLTER can be used to carry out noninvasive prognostic analysis such as heart rate variability analysis, QT dispersion.

Therefore, the ambulatory electrocardiogram of non-volatile recording (HOLTER) in the monitoring of cardiovascular patient, has the irreplaceable advantage of phone transmission electrocardiogram of general electrocardiograph and short record in the diagnosis of cardiovascular disease, be the diagnosis of cardiovascular disease, the important means of the monitoring of cardiovascular patient.But only the HOLTER under broadband networks is supported just might make HOLTER be transferred to work station fast, can monitor everyday, and monitoring becomes health product constantly.

Value of the present invention also is, makes the HOLTER technology enter into infirmary and rural health unit from large hospital.More the patient of wide geographic area can enjoy this inspection.

" national Development of Health Service situation statistical communique in 1999 " according to the issue of Ministry of Public Health health statistics information centre, China's county hospital has 2040, the commune hospital has 5.03 ten thousand (wherein 563 villages and towns in rural areas commune hospitals, avenue commune hospital have 4.97 ten thousand), 4541 of out-patient departments, the clinic, the healthcare center, 22.20 ten thousand of infirmary, amount to 27.8881 ten thousand basic health mechanisms, but average per thousand population doctor numbers have only 1.67 people, still unequipped mostly HOLTER of Duo medical institutions so, even if equipment HOLTER is the few doctor's (can see that especially Electrocardiographic doctor still less) of number.But with possible, solve 27.8881 ten thousand basic health mechanisms with limited resources of human talents and use the HOLTER problem with phone transmission HOLTER.

1999, the cardiopathic mortality rate in urban area was 98.98/10 ten thousand, and the cardiopathic mortality rate in rural area is 74.99/10 ten thousand.Therefore, it approximately is 2086920 people that China died from cardiopathic population in 1999, and the people who suffers from a heart complaint approximately is ten times of death toll, i.e. more than 2,000 ten thousand people.Say that by reason these people should be HOLTER and check, do not have phone transmission HOLTER not realize.Therefore, phone transmission HOLTER has the large market.

Be illustrated below in conjunction with accompanying drawing, the features and advantages of the present invention will become more apparent.

Fig. 1 broadband HOLTER system block diagram

Fig. 2 HOLTER recording box block diagram

The algorithm flow chart of the automatic analyzing ecg of Fig. 3 HOLTER

Fig. 4 low-frequency filter frequency characteristic

Fig. 5 band elimination filter

Fig. 6 50HZ band elimination filter frequecy characteristic

The frequency characteristic of Fig. 7 difference filter

Fig. 8 ECG and differentiated waveform thereof

Choosing of Fig. 9 interpolation point

Figure 10 removes the ecg curve that base floats front and back

Figure 11 detects the P ripple with area-method

Describe in detail below in conjunction with accompanying drawing:

Fig. 1 .1 is one of broadband networks HOLTER system schema, HLOTER recorder (1) becomes digital signal with electrocardiosignal, through radio modem (2), and radiofrequency launcher (3), be transmitted into the wireless receiver (4) in 100 meters scopes of diameter, again behind radio modem (5), the reduction digital signal is delivered to PC computer (6), the latter is through ADSL Modem (7), through the telephone exchange (9) of telephone network (8) to the telephone office, after being coupled to broadband networks (10), through ADSL Modem (11), to work station PC computer (12).

Fig. 1 .2 is two of a broadband networks HOLTER system schema, HOLTER recorder (1), and through modem (2), telephone network (3), telephone exchange (4), broadband networks (5), modem (6) reach the PC computer (7) in the electro cardio signal workstation

Fig. 2 is the block diagram of HOLTER recording box.

According to present practical situation, portable, microminiaturized, low-power consumption product with network data transmission function are made up of analog electrocardiogram signal amplification circuit (I), single-chip microcomputer (CPU) control circuit (II), data transmission circuit (III) and power circuit four parts such as (IV), and this tetrameric major function specifies as follows:

Analog electrocardiogram signal amplification circuit (I)

Impedance matching and channel to channel adapter (1) couple together electrode and amplifier (2).Be used for pick off is carried out processing and amplifying from the faint electro-physiological signals that human body obtains on one's body, also need to be equipped with corresponding A/D translation circuit.I and II are partly coupled together, be about to amplifier and CPU (4) and be coupled together.

Cpu control circuit (II)

According to practical situation, also need to be equipped with control corresponding button (9), condition indication circuit (8) usually for storing the data storage circuitry that electrocardiogram (ECG) data is used, adopt Flash memorizer (7) usually, produce the RTC circuit (6) of real-time clock signal etc.

Data transmission circuit (III)

After the circuit of above-mentioned two parts was finished amplification, collection and the memory function of electrocardiosignal, the transport process of data was just finished by data transmission circuit (5).By radio frequency transmitter (11) signal is transmitted into the space, receives by the radio frequency receiver among Fig. 1 (4).

Power supply circuits (IV)

Because portable, microminiaturized equipment uses battery powered usually, therefore need to use the power management device to finish necessary voltage stabilizing and voltage transformation (10).What power supply circuits need be considered usually always has: output voltage and available maximum current must satisfy that actual needs (electric current also need keep suitable surplus), power conversion efficient are high as far as possible, the required power consumption of power supply circuits itself is low as far as possible, preferably have battery low pressure measuring ability, circuit realizes that simple (peripheral component lacks as far as possible), cost are suitable etc.

Fig. 3 is the algorithm flow chart of the automatic analyzing ecg of HOLTER, is described in detail as follows:

" quality " of ECG quality is meant that ECG is subjected to the degree of noise jamming, and the general source of these noises is:

Because the low-frequency disturbance (being baseline drift) that breathing, loose contact or patient posture are moved and amplifier etc. brought

Because the power frequency (50HZ) that power line brought is disturbed

Because the High-frequency Interference that myoelectricity produced

The design of the S04 low pass filter among Fig. 3

Realize with assembler language in order to improve arithmetic speed and to be convenient to, the wave filter that adopts all is the integral coefficient digital filter, it can be in time domain difference equation Recursive Implementation, also can use rapid fourier change (FFT) to realize at frequency domain, more helps improving arithmetic speed but implement in time domain.

The effective frequency composition of electrocardiosignal is distributed in 0 to 44HZ, and the transfer function of general integral coefficient low pass filter is $H\left(z\right)={\left(\frac{1-b^{-m}}{1-z^{-1}}\right)}^k---\left(2.1\right)$

The attenuation characteristic of K decision wave filter in the formula, when K=2, the difference of M is chosen the 3dB frequency that has determined wave filter, and its relation is as follows: m:2 345678 93db frequently: 90.9 56.0 41.0 32.5 26.9 23 20 17.8hz

Through relatively, get K=2, M=5.This moment 3dB frequency f c=32.5hz, the gain at 100hz place is zero, can eliminate 2 times of interference of power frequency, transfer function at this moment: $H\left(Z\right)={\left(\frac{1-Z^{-5}}{1-Z^{-1}}\right)}^2=\frac{1-2Z^{-5}+Z^{-10}}{1-2Z^{-1}+Z^{-2}}---\left(2.2\right)$ Its frequency characteristic $H\left(e^{\mathrm{jw}}\right)=\frac{\mathrm{<figure-callout\; id="2"\; label="si\; n"\; filenames="A0114359200171.png"\; state="\{\{state\}\}">si</figure-callout>}{n}^{}{}^2\frac{5}{2}\mathrm{\&omega;}}{{\mathrm{<figure-callout\; id="2"\; label="sin"\; filenames="A0114359200171.png"\; state="\{\{state\}\}">sin</figure-callout>}}^2\frac{1}{2}\mathrm{\&omega;}}{e}^{-\mathrm{Aw}}---\left(2.3\right)$ Obviously it has linear phase, and we are desired just for this.Corresponding time domain equation is: y (n)=x (n)-2x (n-5)+x (n-10)+2y (n-1)-its frequency characteristic of y (n-2) (2.4) is shown in Figure 4

S03 among Fig. 3 eliminates the design of power frequency interference-50HZ band-rejection digital filter

Because directly design band resistance filter is relatively more difficult, therefore, we then adopt that to design earlier an all-pass filter and one respectively be the band filter of mid frequency with 50HZ, the two is subtracted each other, can obtain one is the band elimination filter of mid frequency with 50HZ, as shown in Figure 5.

We are at designed 50HZ band elimination filter transfer function: $H\left(z\right)={5z}^{-10}-\frac{1+z^{-25}}{1+z^{-5}}---\left(2.5\right)$

Domain equation and frequency domain equation are omitted at that time, its frequency characteristic as shown in Figure 6, as seen it has the stopband that is rather narrow, and has effectively suppressed the 50HZ interference, and don't influences cardiac electrical feature.

Low-frequency disturbance, that is the removal of baseline drift mainly are the detections that is used for the P ripple, and therefore, this a part of content is put into the P ripple and detects narration in the joint.

The ecg characteristics amount detects accurately, it is the basis of realizing automatic diagnosis, detected characteristic quantity is many more, come the accuracy and the probability of automatic diagnosis big more, but also make the algorithm complexity, to real-time implementation band difficulty, in this project, the main foundation that sick class is differentiated algorithm is the width of R-R interval and QRS wave group.Therefore, effectively detecting these two parameters promptly is to finish the key of this task.

To determine accurately that R-R at interval, at first to detect the QRS wave group exactly, promptly give R ripple location, the detection of R ripple be all computer electrocardio automatic analysis systems the common issue with that must solve, but regrettably, so far go back the neither one algorithm and can accomplish that 100% R ripple detects, this mainly is because the variation and the noise jamming of R ripple attitude.

The detection of the S06R ripple of Fig. 3 of the present invention

The characteristics of R ripple are to have a comparatively tall and big spike (mostly be forward, also may oppositely) and the variation of current potential very fast, when therefore detecting, can detect based on amplitude, detect but more be based on its slope, and this task promptly is to adopt the slope method to detect the R ripple.

A digital signal x (n), its slope can be obtained by following formula:

y(n)＝(Δ1+Δ2+Δ3+Δ4)/6 (2.6)

Δ k=x (m+k) in the formula-x (m-k) k=1,2,3,4

Its amplitude-frequency characteristic is $\left|H\right(e^{\mathrm{jw}})|=8\mathrm{<figure-callout\; id="5"\; label="Sin"\; filenames="A0114359200171.png"\; state="\{\{state\}\}">Sin</figure-callout>}\frac{5}{2}\mathrm{\&omega;}\&CenterDot;\cos \mathrm{\&omega;}\&CenterDot;\cos \frac{\mathrm{\&omega;}}{2}---\left(2.7\right)$ It is similar to the frequency characteristic jw of desirable difference filter in low-frequency range, as shown in Figure 7.

And (2.6) formula is a coefficient is 1 FIR wave filter entirely, and it is extremely easy to implement.Like this, a QRS wave group through differentiated waveform as shown in Figure 8.

The step that the QRS wave group of the S06 of Fig. 3 of the present invention detects is as follows:

(1) the raw ECG waveform is made calculus of differences, obtain Fig. 8 (b)

(2) choose the certain proportion of 8 peaked meansigma methodss of heartbeat difference as detection threshold T _KRShown in figure (8b).

(3) when the difference value of this sample point during, think to begin to enter the R ripple, remember that this point is C greater than threshold value.

(4) when the difference value of back during, think temporarily and leave the R ripple, be designated as the D1 point greater than threshold value, if between the C-D1 every too narrow, continue to look for one less than threshold point D2, if the C-D2 distance thinks then and find a R ripple that the C-D2 distance is thought the width of R ripple in allowed band.

(5) whether if do not find the R ripple within greater than 1.5S, then reduce threshold value, remaking binary search backward has amplitude to become less R ripple suddenly to search.

Owing to taked above measure, therefore can determine the R ripple of various forms exactly, detect and reach 98%.

The detection of the QRS width of Fig. 3 S08 of the present invention

Shown in Fig. 8 (b)

(1) after detecting a R ripple, search for backward by the C point, when continuous 8 points continuously less than threshold value T _KRThe time, thinking the onset point that finds Q, Qon is the width of Q ripple to the distance of Qon-c.

(2) search for forward by D1 or D2 point, when continuous 8 points less than threshold value T _KRThe time, think terminal point (offset) S that finds the S ripple _Of, then the distance of D2-Sof is thought the width of S ripple.

(3) width of Qon-Sof is (a Q wave width+R wave width+S wave width), and it promptly is the width of QRS ripple.

(4) when detection width, in order to prevent noise jamming, we have taked current heartbeat width constantly and preceding 4 methods that equal width compares of clapping, and with the effectiveness of determining that this bats detects, even the width of this bat obviously thinks just that greater than the mean breadths of preceding 4 bats and greater than 120ms it is broadening.

(5) detect at the R ripple, in the testing process that Qon and Sof are ordered, threshold value T _KP, T _KQ, T _RSObtain by the self study process, and bring in constant renewal in, to adapt to the cardiac electrical variation of patient.

Algorithm is removed in the baseline drift of the S07 of Fig. 3:

Because baseline drift belongs to low-frequency disturbance.Therefore, the method of going base to float commonly used at present has high-pass filtering method, and linear correction method and cubic spline function correction method are because high-pass filtering method can make the distortion of ST section again when removing basic floating, therefore reproached, the amount of calculation of linear correction method is few, and is easy to use, but correction mass is lower, the cubic spline function correction method can obtain higher correction accuracy, but the algorithm complexity, operand is big, can't real-time implementation.

At not only requiring the quality height but also requiring characteristics of real time, the present invention proposes a kind of cubic curve fitting process, promptly choose interpolation point Ij in the T-P section, choose interpolation point I in next Tp section _J+1, as shown in Figure 9.

Cross these two interpolation points and make a cubic curve

y＝ax ³+bx ²+cx+d

Come match IJ to I _J+1Baseline, utilize two interpolation points functional value (y1, y2) and derivative value (y1 y2) can obtain four coefficient a, b, c, d $a=\frac{{\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="A0114359200171.png,A0114359200213.png"\; state="\{\{state\}\}">y</figure-callout>}}_1^{\&prime;}+{\mathrm{<figure-callout\; id="2"\; label="y"\; filenames="A0114359200171.png"\; state="\{\{state\}\}">y</figure-callout>}}_2^{\&prime;}-2y^{\&prime;M}}{{(x_2-x_1)}^2}---\left(2.8a\right)$ $b=\frac{y_{\mathrm{ax}}^{\&prime;}-{\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="A0114359200171.png,A0114359200213.png"\; state="\{\{state\}\}">y</figure-callout>}}_1^{\&prime;}}{x_2-x_1}-a(x_2+2x_1)---\left(2.8b\right)$ $c={\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="A0114359200171.png,A0114359200213.png"\; state="\{\{state\}\}">y</figure-callout>}}_1^{\&prime;}-(3{\mathrm{<figure-callout\; id="1"\; label="ax"\; filenames="A0114359200171.png,A0114359200213.png"\; state="\{\{state\}\}">ax</figure-callout>}}_1^2+3b{x}_1)---\left(2.8c\right)$ $d=y_1-({\mathrm{<figure-callout\; id="1"\; label="ax"\; filenames="A0114359200171.png,A0114359200213.png"\; state="\{\{state\}\}">ax</figure-callout>}}_1^3+{\mathrm{<figure-callout\; id="1"\; label="bx"\; filenames="A0114359200171.png,A0114359200213.png"\; state="\{\{state\}\}">bx</figure-callout>}}_1^2+{\mathrm{cx}}_1)---\left(2.8d\right)$

In the formula $y_{\mathrm{ax}}^{\&prime;}=\frac{y_2-y_1}{x_2-x_1}---\left(2.8e\right)$

Obtain curve y=ax behind a, b, c, the d ³+ bx ³+ cx+d can constitute like this, can eliminate base by following formula and float:

y _out＝y _in-C(ax+b)·x+cx+d X ₁≤X≤X ₂ (2.9)

Y in the formula _InBe original electrocardiographicdigital data, y _OutFor removing the data after base floats

It is shown in Figure 10 utilizing this algorithm to remove the example that base floats

When this algorithm of research, we have at length inquired into the principle that interpolation point is chosen, and have proposed a fitting of parabola correction method on this basis, have further improved arithmetic speed.

The detection of the P ripple of Fig. 3 S10 of the present invention

The present invention has at first proposed a kind of dual threshold P ripple based on slope and has detected and location algorithm, and a bigger threshold value is used for the P ripple and detects, and another less threshold value is used for locating P ripple starting point and terminal point, and specific practice is:

(1) raw ECG is done pretreatment and baseline correction

(2) be decided to be Q, R, T ripple, with the region of search of T-Q interval as the P ripple

(3) further eliminate myoelectricity with the digital filter that is generally 0-20HZ and disturb (Fig. 3

S09).

(4) the T-Q section is done five and order the parabola differentiate, find out its derivative maximum, to continuously

5 such maximums are asked the average average maximum a of getting _Ym

(5) utilize a _YmWith getting two adaptive threshold th _PcAnd th _p

Th _PcBe used for detecting the P ripple, th _pBe used for locating Ponset and Poffset, they are all set up through self study, and applicable to different patient's electrocardiosignaies, 1.0 in the following formula and 0.25 is to consider that the baseline after the correction is not absolute straight line, and be clamped for preventing the wide lower limit that adds of criterion.

Make th _PcAnd th _pAfterwards, moving back back search from the Qonset point, till the derivative absolute value of continuous 4 points was greater than thpc, this moment, the P ripple was detected, and had not found that yet continuous 4 derivative absolute values are greater than thpc if search Toffset.Think that then this heartbeat does not have the P ripple.

After detecting the P ripple, be detected a little from the P ripple, utilize threshold value thp forward, search for backward, make the starting point and the terminal point of P ripple, so far the P ripple detects and finishes.We have made practice examining to above-mentioned algorithm,

For the existence of the more effective definite P ripple of energy and the position of terminus thereof, the present invention is on the basis of above-mentioned two-way threshold method, proposed to detect the new method of P ripple again with the area method of addition, its basic thought is: P corrugated as shown in figure 11 is long-pending, can regard as many leg-of-mutton areas and, peak value with the P ripple is the center, and its area is a bit calculated in each extension to both sides, and then extends a point in one direction again.Calculate its area increment, pointwise is calculated down, and when the difference of double area was zero, the point on the peak value left side was P ripple starting point Ponset, and the point on peak value the right is terminal point Poffset.

Each area increment by following formula calculated curve X (i)

2S ₁＝{X(i)-X(i-1)}{i+1-(i-1)}

2S ₂＝{X(i+1)-X(i+2)}{i+2-(i-1)}

2S ₃＝{X(i-1)-X(i-2)}{+2-(i-1)}

2S ₄＝{X(i+2)-X(i+3)}{i+3-(i-2)}

2S ₅＝{X(i-2)-X(i-3)}{i+3-(i-3)}

SP=S1+S2+S3+S4 is amassed on the P corrugated

With this execution follow procedure section that circulates, can realize that the P ripple detects

SS(K-1)＝K|X(i-1)-X(i-1-1)|

SS(K)＝(K+1)|X(i+1+1)·X(i+2+1)

S＝SS(K-1)+SS(K)+S

K＝K+2

Wherein: the initial value of K is 2, and SS (K-1) is the area increment, and 1 is cyclic variable, and 1 is peak point, and X (n) is the amplitude of ecg wave form.

See with regard to actual result of the present invention, to range estimation can see real storage really the P ripple, inspection mountain rate can reach more than 82%, in this seen at home report than higher.

So far, we have narrated this analysis instrument in more detail from signals collecting, and pretreatment is to the ultimate principle of parameter detecting and the key issue that is solved.

The sick class method of discrimination of the S11 of Fig. 3 of the present invention:

The employed parameter of following criterion is: interval instantaneous and average R-R interval, chamber rate, room rate (PP interval), instantaneous and average QRS width, PR interval, mean P-P etc., corresponding symbol are:

R-R interval between RR (I) I bat and I-1 clap

AR (I): the average R-R interval that the normal heart of clapping corresponding to I is clapped

ARI (I): corresponding to the average R-R interval of I bat

The QRS width that DQRS (I) claps corresponding to I

AD (I): the meansigma methods that the heart of the normal QRS width of clapping corresponding to I is clapped

The P-R interval that PI (I) claps corresponding to I

PR (I) claps the P-R interval corresponding to I

The meansigma methods of the P-P interval of the normal P ripple that AP (I) claps corresponding to I

The criterion of 15 kinds of arrhythmias will differentiating the present invention is:

(1). stop fighting

Sinus arrest and ventricular asystole are incorporated in-rise, as long as do not find the QRS wave group in 3 seconds, promptly RR (I)＞3S then thinks to stop fighting

(2). leak and fight (comprising second degree A-V block)

RR(I)＜3.2AR(i)

RR(I)＞1.7AR(i)

In time, thought to leak and fights

(3). tachycardia

Continuous 5 heartbeats above R-R intervals, think that less than 3/7 second or during greater than the heart rate threshold that is provided with this section electrocardiosignal is a tachycardia.

(4). bradycardia

Continuous R-R more than 5 seconds at interval greater than 1.5 seconds or during less than the heart rate threshold that is provided with, thinks that this section electrocardiosignal is a bradycardia.

(5). the chamber early:

Satisfy following condition:

1. the QRS wave group broadening of I-1 bat

2. but I claps full remuneration

3. I-2 bat width is normal

4. before the QRS wave group phase that I-1 claps

5. the QRS width of I bat is normal

Think that then it is a ventricular premature contraction that I-1 claps

(6). ventricular bigeminy condition: 1. before the QRS wave group phase that I-1, I-3 two clap;

2. clap can full remuneration for I, I-2 two;

3. the QRS burst width of I-1, I-3 two bats is normal

4. the QRS burst width broadening of I, I-2, I-4 triple time

Promptly find two successive ventricular premature contractions, then think a premature ventricular beat to have occurred

(7). chamber property premature ventricular beat condition: 1. before the QRS wave group phase that I-1, I-4 two clap.

2. clap can full remuneration for I, I-3 two.

3. the QRS burst width broadening of I-1, I-4 two bats

4. I, I-2, I-3, I-5, each QRS wave width of clapping of I-6 normally satisfy above condition, then think a premature ventricular beat to have occurred.

(8). interpolated premature beat (can be chamber property, also can be boundary property) condition: 1. to clap the QRS wave width normal for I, I-2

2. I-1 claps broadening

3. I claps does not have the amortization period

(9) .RonT condition: 1. RR (I)＜RR (I-1) of clapping of I

2. before the I-1 bat phase, promptly AR (I)＞1 second and RR (I)＜1/3S or ARI (I) 〉=0.6S and ARI (I)＜IS and RR (I)≤AR (I)

(10). running fire ventricular premature contraction condition: 1. I-1, I-2 two clapped the QRS ripple before the phase

2. I-1, I-2 two clap the broadening of QRS wave width

3. the QRS wave width of I, I-3 two bats is normal

(11) .-degree atrioventricular block

In 16S, there is the heart more than 2/3 to clap to meet the following conditions and then is judged to the atrioventricular block of I degree

①PR(I)＞0.22S

2. I bat QRS wave width is normal

(12). supraventricular premature beat (comprise room early, junctional premature beat) condition: 1. I, I-1, that I-2 claps the QRS wave width is normal

2. before the I-1 bat phase, RR (I-1)＜0.9AR (I-1) is arranged

3. supraventricular premature beat is then thought in I bat full remuneration

(13). the room is condition early: 1. I-1 claps the P ripple in advance, and PI (I-1)＜0.9AP (I-1) is arranged

2. to clap the QRS wave width normal for I, I-1, I-2

3. I claps the imperfection compensation

(14). the chamber quiver 1. first three greatest gradient of ECG waveform average absolute less than 200 2. the difference of crest, the trough number of ECG waveform be less than or equal to mean number 30% 3. the ECG waveform the number of oscillation greater than 200 times/minute 4. the meansigma methods of crest, the trough amplitude of ECG less than 20% of average amplitude

(15). the chamber of taking place frequently is early

Per minute chamber number morning is greater than 5.

Claims (7)

1. broadband HOLTER system, comprise HOLTER recorder (1), radio modem (2), radiofrequency launcher (3), radio frequency receiver (4), radio modem (5), PC (6), ADSL Modem (7), telephone network (8), telephone exchange (9), broadband networks (10) ADSL Modem (11), PC computer (12).

2. according to the system of claim 1, it is characterized in that: described PC computer (12) is an ECA, and it comprises the signal conditioner of a low-frequency filter, and its transfer function is: $H\left(Z\right)={\left(\frac{1-b^{-m}}{1-Z^{-1}}\right)}^k$

3. according to the system of claim 1, it is characterized in that: described PC computer (12) is an ECA, and it comprises a 50HZ band elimination filter, and the transfer function of this device is: $H\left(Z\right)={5Z}^{-10}-\frac{{1+Z}^{-25}}{{1+Z}^{-5}}$

4. according to the system of claim 1, it is characterized in that: described PC computer (12) is an ECA, and it comprises a QRS ripple checkout gear, and its amplitude-frequency characteristic is: $\left|{\mathrm{He}}^{\mathrm{jw}}\right|=8\mathrm{Sin}\frac{5}{2}\mathrm{\&omega;}\&CenterDot;\cos \mathrm{\&omega;}\&CenterDot;\cos \frac{\mathrm{\&omega;}}{2}$

5. according to the system of claim 1, it is characterized in that: described PC computer (12) is an ECA, and it comprises that is removed the pretreatment unit that base floats, and this device is eliminated base by following algorithm and floated: $y_{\mathrm{out}}=y_{\mathrm{in}}-C(\mathrm{ax}+b)x+\mathrm{cx}+d.$

6. according to the system of claim 1, it is characterized in that: described PC computer (12) is an ECA, and it comprises a P ripple checkout gear, and this device is to detect P ripple and location with area method of addition and dual threshold method.

7. according to the system of claim 1, it is characterized in that: described PC computer (12) is an ECA, it comprises a sick class discriminating gear, it can determine stop fighting, leak fight, tachycardia, bradycardia, chamber morning, ventricular bigeminy, chamber property premature ventricular beat, interpolated premature beat, RONT, running fire ventricular premature contraction, I ° degree atrioventricular block, supraventricular premature beat, room early, the chamber quivers, the chamber of taking place frequently early.

Images