CN100515335C - Blood oxygen measuring method and device capable of eliminating moving inteference - Google Patents

Abstract

The present invention is blood oxygen measuring method and device capable of eliminating motion interference. The blood oxygen measuring method includes the following steps: converting the light signal of pulse wave and blood oxygen saturation degree information into electric signal, amplifying the electric signal, A/D converting, data processing and calculating in the monolithic computer; developing the spectrum of the data in the algorithm module and converting the motion noise of the blood oxygen waveform into one constant; differential operation to filter out the motion noise and normalizing treatment to filter off the base line drift noise; filtering out the step noise and pulse noise via centering; and final area integral operation to reduce the pulse wave and blood oxygen waveform. The method and device can eliminate the influence of the testee's finger motion on the measurement basically, and the measurement is reliable and accurate.

Description

Can eliminate the blood oxygen measuring method and the device thereof of motion artifacts

Technical field

The present invention relates to measure the method and apparatus of the heart and blood oxygen, relate in particular to utilization sampled signal is carried out the signal conversion to get rid of the measuring method and the device of interferometry precision owing to move.

Background technology

The basic structure of oxygen saturation measurement device comprises blood oxygen probe and signal processing apparatus both at home and abroad at present, and blood oxygen probe is a pick off that adopts light emitting diode and light-sensitive element to form.Light emitting diode provides the light of two or more wavelength.The effect of light-sensitive element is that handle converts the signal of telecommunication to by organizing the terminal optical signal that has blood oxygen saturation information, and this is signal digitalized.For this digital quantity, the blood oxygen saturation that adopts certain signal processing algorithm to calculate.

Because when patient is carried out oxygen saturation monitor; patients fingers (or toe) moves through regular meeting; distance between finger (or toe) and the pick off is subjected to displacement; thereby cause patient's pulse wave of measuring very unstable; signal to noise ratio is very low; adopt existing method to measure blood oxygen; need to measure the alternating component of pulse wave; just to seek maximum and minima in the pulse wave; and often cause under the motion conditions mistake of crest in the pulse wave and trough is sought; be difficult to obtain being worth most accurately, thereby make that the Oximetry precision of this moment is very poor.

U.S. Pat 4,714,341 disclose the technology that a key name is called " the multi-wavelength oximeter that need not consider interfering signal ", this technology is for example finger with the light transmission human body of three beams different wave length, finding out thus obtained three groups contains the blood oxygen saturation parameter and has difference between the perceptual signal of different wave length, judge whether they drop in the preset range, thereby interference factor is excluded.This public technology hardware configuration complexity, the software computing is loaded down with trivial details, and certainty of measurement is still not really desirable.Another U.S. Pat 5,482,036 discloses the technology that a key name is called the measurement blood oxygen saturation of " signal processing apparatus and method " again, this technology utilizes a kind of signal processor to produce the noise contrast signal of ingenious structure, by a module that is called as the self adaptation noise killer, from measure each signal of gained, remove undesired part again.Described self adaptation noise killer comprises a joint treatment estimation device, and the latter comprises a minimum square dot matrix predictor and a recurrence wave filter again.The transfer function of this technology and soft, hardware configuration are still too complicated, cost is high, and it is also so desirable to remove noise result.

Summary of the invention

The technical problem to be solved in the present invention is to avoid above-mentioned the deficiencies in the prior art part and proposes a kind of can effectively the minimizing because Oximetry data processing method and the oxygen saturation measurement device that motion exerts an influence to certainty of measurement.

The present invention adopts following technical scheme: propose a kind of Oximetry data processing method of eliminating motion artifacts, this method comprises the steps:

A. probe with detected have optical signal that pulse involves blood oxygen saturation information and be converted to the signal of telecommunication after, again this signal of telecommunication is carried out processing and amplifying, then this signal of telecommunication is admitted to the A/D module;

The b.A/D modular converter is the analog signal conversion through processing and amplifying a digital signal;

The digital signal of c.A/D modular converter output is admitted to single-chip microcomputer and carries out date processing and calculating;

Especially, date processing and the calculating that described single-chip microcomputer carried out comprises the steps:

D. pulse wave and blood oxygen waveform are regarded as the stack of certain waveform and noise, thus the HONGGUANG of reflection blood oxygen waveform characteristic and the time series of infrared light pulse wave signal are expressed as the synthetic of different frequency range wave spectrum, promptly the time series of initial data being carried out wave spectrum launches, in the wave spectrum space, the motion noise of blood oxygen waveform is for conversion into a constant;

E. the wave spectrum expression formula of above-mentioned initial data being differentiated, is zero behind the motion noise differential as constant, has promptly filtered motion noise by differentiating;

F. the slow baseline drift of initial data shows as an oblique line in wave amplitude time coordinate system, has the waveform of physical significance to be superimposed upon on this oblique line, filters the baseline drift noise by normalized

G. the noise that shake is caused shows as the pulse noise of step noise or delta-function in wave amplitude time coordinate system, the pulse noise of this step noise and delta-function is got intermediate value and adjacent 5 data points by adjacent 3 data points respectively and got intermediate value and be filtered;

H. by the HONGGUANG of having eliminated baseline drift, step noise and pulse noise in the wave spectrum space and infrared light pulse wave signal being carried out described pulse wave and blood oxygen waveform are reduced in the area integral computing of sampling period;

I. the forgetting factor λ that introduces value between 0 and 1 eliminates the data saturated phenomenon of long-time integration, promptly to measure constantly and the result of calculation of previous reasonable period carry out the result of calculation that iteration obtains this measurements moment, the approaching more described data constantly of measuring are big more to the contribution of result of calculation during iteration;

J. the pulse wave of step I acquisition and blood oxygen saturation are calculated the result by outwards exporting with the serial ports of described single-chip microcomputer.

The described unfolded function of wave spectrum that is used for is other functions that sinusoidal wave function, cosine wave function maybe can constitute complete orthogonal system.The integration period of described step h is 2 to 3 seconds; Described forgetting factor λ value is 0.8, and the iterative relation formula of introducing this forgetting factor λ is as follows:

$\frac{{\mathrm{Red}}_{\mathrm{AC}}}{{\mathrm{Ir}}_{\mathrm{AC}}}=\frac{{\mathrm{Red}}_{{\mathrm{AC}}_0}+\mathrm{\λ}{\mathrm{Red}}_{{\mathrm{AC}}_1}+\·\·\·+{\mathrm{\λ}}^{n}R{\mathrm{ed}}_{{\mathrm{AC}}_n}}{{\mathrm{Ir}}_{{\mathrm{AC}}_0}+\mathrm{\λ}{\mathrm{Ir}}_{{\mathrm{AC}}_1}+\·\·\·+{\mathrm{\λ}}^n{\mathrm{Ir}}_{{\mathrm{AC}}_n}}.$

Described probe utilizes the two-way light transmission to organize end to carry out sampled measurements; Wherein one the tunnel is HONGGUANG, and corresponding data integrate result is that the red light intensity that receives in this section period exchanges peak-to-peak value, and another road is an infrared light, and corresponding data integrate result is that the infrared light intensity that receives in this section period exchanges peak-to-peak value.

Comprised also before step a that system powers on, the sampling set-up procedure of hardware initialization, cpu system self check and program initialization.

The technical solution used in the present invention also comprises:

A kind of blood oxygen measuring device that can eliminate motion artifacts, comprise: the probe assembly that is linked in sequence, signal processing and amplifying assembly, A/D conversion, single-chip data processing components, Logic control module and power circuit, it is characterized in that: described single-chip data processing components comprises the motion artifacts cancellation module, and the outfan of this motion artifacts cancellation module connects A/D conversion and logic control and functional module respectively.

Described motion artifacts cancellation module comprises pulse wave calculation of parameter, differential processing module, medium filtering module, integration recovery module and the area integral recurrence module of the unidirectional connection of order, and the output result of this area integral recurrence module is used to calculate pulse wave and blood oxygen saturation.

Described probe assembly comprises HONGGUANG, infrared light supply and light drive circuit thereof and light-receiving tube.Described signal enlargement processing section branch comprises differential amplifier circuit and signal processing circuit.

Described logic control and functional module comprise security function and communication function; Described global function comprises by pick off and comes off, searches for pulse, the oversize and blood oxygen saturation of search pulse descend measurement Status Flag of being formed and the module self check of finishing error report; Affiliated communication function comprises receiving orders and sends data.

Compared with prior art, blood oxygen measuring method and device thereof that the present invention can eliminate motion artifacts have following advantage: adopt the pulse signal differential is carried out the albefaction processing, increased the reliability that pulse wave is worth searching most.And then the pulse signal area integral replaced in the traditional method searching to waveform peak.Do not need each pulse wave is searched out accurately value, only need carry out integration the pulse wave of a period of time.Because the integration of noise goes to zero in a period of time, can remove the influence of noise to the integration of pulse wave, can prove that the result of pulse wave integration is equal to the alternating component of pulse wave, thereby can utilize integral result to calculate blood oxygen saturation, the result of calculation of this blood oxygen saturation numerical value has filtered basically because patients fingers is moved to the influence of measurement result.

Description of drawings

Fig. 1 is the functional-block diagram that the present invention can eliminate the blood oxygen measuring device of motion artifacts;

Fig. 2 is reduced hemoglobin and the HbO2 Oxyhemoglobin change curve at the absorption coefficient of light in HONGGUANG and infrared light district;

Fig. 3 is the extinction sketch map of measured tissue;

Fig. 4 is the pulse waveform that comprises noise;

Fig. 5 is the pulse waveform figure that comprises baseline drift;

Fig. 6 is the pulse waveform figure that eliminates baseline drift;

Fig. 7 is the pulse waveform figure that comprises the step noise;

Fig. 8 is the pulse waveform figure that comprises the step noise behind the differential;

Fig. 9 is the pulse waveform figure that comprises pulse noise;

Figure 10 is the pulse waveform figure that comprises pulse noise behind the differential;

Figure 11 is the pulse waveform figure corresponding with Fig. 4 that comprises noise behind the differential;

Figure 12 is a pulse waveform area integral oscillogram;

Figure 13 is the pie graph of motion noise cancellation module and functional module.

The specific embodiment

Below in conjunction with the embodiment shown in accompanying drawing and the accompanying drawing apparatus of the present invention and method are described in further detail.

As shown in Figure 1, a kind of blood oxygen measuring device that can eliminate motion artifacts of the present invention comprises: the probe assembly that is linked in sequence, signal processing and amplifying assembly, A/D conversion, single-chip data processing components, Logic control module and power circuit.The crucial part that apparatus of the present invention are different from prior art is that described single-chip data processing components has comprised the motion artifacts cancellation module, and the outfan of this motion artifacts cancellation module connects A/D conversion and logic control and functional module respectively.

Described probe assembly comprises HONGGUANG, infrared light supply and light drive circuit thereof and light-receiving tube.Described signal enlargement processing section branch comprises differential amplifier circuit and signal processing circuit.

Figure 13 is the pie graph of motion noise cancellation module and functional module.Described motion artifacts cancellation module comprises pulse wave calculation of parameter, differential processing module, the medium filtering module of the unidirectional connection of order, the area integral recurrence module of integration recovery module, and the output result of this area integral recurrence module is used to calculate pulse wave and blood oxygen saturation.Described logic control and functional module comprise security function and communication function; Described global function comprises by pick off and comes off, searches for pulse, the oversize and blood oxygen saturation of search pulse descend measurement Status Flag of being formed and the module self check of finishing error report; Affiliated communication function comprises receiving orders and sends data.

The function of each several part is as follows:

Power circuit is input as ± 12V AC and DC power supply, exports two groups of power supplys---numeral+5V, simulation ± 5V, and this circuit provides power supply for whole blood oxygen plate.

The photoelectricity drive circuit is subjected to the adjusting of logic control part, exports the current drives light emitting diode of different amplitudes, can export the signal of certain amplitude to guarantee light-receiving tube.

Probe is sent into signal processing and amplifying part after detected optical signal is converted to the signal of telecommunication, through difference amplify, the bias light electric current is wiped out processing, gain-adjusted, bias current are wiped out processing, send A/D conversion at last.

The A/D conversion portion is a digital signal with the analog signal conversion after processing and amplifying, send single-chip microcomputer to handle then.

The date processing that the single-chip data processing section is converted to A/D, calculating obtain pulse wave and blood oxygen saturation.

Pulse wave and blood oxygen saturation data also outwards send by serial ports, and have optocoupler to be isolated.

In addition, single-chip microcomputer partly realizes by logic control the control of various piece, as the luminous sequencing contro of popping one's head in, drive current control, bias current control, bias light are wiped out control, signal A/D conversion and control etc.

The operation principle of brief description apparatus of the present invention.

Pulse formula oximeter is to utilize beating of arterial blood in the tissue end that pulse wave causes, because HbO2 Oxyhemoglobin is different in the optical characteristics of red-light spectrum district and ultrared with reduced hemoglobin, thereby the influence finger is to the light transmission of HONGGUANG and infrared light, when the HONGGUANG of certain light intensity and infrared light are added to finger when going up, by detecting the transmitted light intensity of two kinds of wavelength respectively, can calculate the content of HbO2 Oxyhemoglobin to the ratio meter of two kinds of light optical density variable quantities by finger, thus the blood oxygen saturation of calculating.

Usually, use the metric measurement blood oxygen saturation, light transmission and reflective optical methods are wherein arranged.All, utilize the difference of the absorption coefficient of light of reduced hemoglobin and HbO2 Oxyhemoglobin to carry out based on Lambert-Beer's law and light scattering theory.As shown in Figure 2.Lambert-Beer's law is:

I＝I ₀e ^-εcd

Wherein, I: transmitted light intensity; I ₀: incident intensity; C: be subjected to the stimulative substance solution concentration; D: solution is subjected to optical path length; ε: the light absorption constant of material; Release by following formula:

D＝lnI ₀/I＝εcd

D claims optical density.Formula has been found that the light absorption situation of material and the dependency of its concentration thus, and this is also just indicating the probability of being calculated the organization internal composition by the light absorption situation of tissue.

Research worker have been studied again and blood oxygen saturation closely-related two big component Hb and HbO further ₂Optical absorption characteristics, found both marked difference, shown in the hemoglobin absorption curve of Fig. 2.Can be seen by Fig. 2, be the HbO of visible red place of 660nm at wavelength ₂The absorption coefficient of light only be 1/10 of Hb, at Hb of infrared light place and the HbO of 805nm ₂One isobestic point is arranged, and at the infrared light district of 940nm HbO ₂The absorption coefficient of light then greater than Hb.

SaO ₂＝HbO ₂/(Hb+HbO ₂)＝C ₁/(C ₁+C ₂)(1)

$D\left(660\right)=\ln I_0\left(660\right)/I\left(660\right)=\ln (I_0\left(660\right)/I\left(660\right)e^{-{\mathrm{\ϵ}}_1{c}_{1}d}{e}^{-{\mathrm{\ϵ}}_2{c}_{2}d})={\mathrm{\ϵ}}_1{c}_{1}d+{\mathrm{\ϵ}}_2{c}_{2}d---\left(2\right)$

$D\left(805\right)=\ln I_0\left(805\right)/I\left(805\right)=\ln (I_0\left(805\right)/I\left(805\right)e^{-{\mathrm{\ϵ}}_3{c}_{1}d}{e}^{-{\mathrm{\ϵ}}_4{c}_{2}d})={\mathrm{\ϵ}}_3{c}_{1}d+{\mathrm{\ϵ}}_4{c}_{2}d---\left(3\right)$

Wherein, SaO ₂Be arterial oxygen saturation; C ₁Be HbO ₂Concentration, C ₂Be Hb concentration; I ₀, I is respectively the light intensity of incident illumination and transillumination; ε ₁, ε ₂Be respectively HbO ₂, Hb is to the absorbance of the HONGGUANG of 660nm wavelength; ε ₃, ε ₄Be respectively HbO ₂, Hb is to the absorbance of the infrared light of 805nm wavelength, and ε ₃=ε ₄=ε; D is a transmission tissue thickness; Get by (2), (3):

C ₁+C ₂＝D(805)/εd；

C ₁＝(D(660)-ε ₂D(805)ε)/(ε ₁-ε ₂)d；

Substitution (1):

SaO ₂＝A×D(660)/D(805)+B (4)

Wherein, A=ε/(ε ₁-ε ₂); B=ε ₂/ (ε ₁-ε ₂);

But, D (660), D (805) be not as in the above (2), (3) formula expressed only with Hb and HbO ₂Relevant, but relevant with the absorbing state of muscle, skeleton, pigment, fat, venous blood etc. in the tissue, promptly D (660), D (805) also should comprise a background absorption portion, as shown in Figure 3.Therefore (2), the evolution of (3) formula are:

$D\left(660\right)=\ln I_0\left(660\right)/I\left(660\right)=\ln (I_0\left(660\right)/I_B{e}^{-{\mathrm{\&epsiv;}}_1{\mathrm{<figure-callout\; id="1"\; label="c"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">c</figure-callout>}}_1\mathrm{\&Delta;d}}{e}^{-{\mathrm{\&epsiv;}}_2{c}_2\mathrm{\&Delta;d}})---\left(5\right)$

$D\left(805\right)=\ln I_0\left(805\right)/I\left(805\right)=\ln (I_0\left(805\right)/I_B{e}^{-{\mathrm{\&epsiv;}}_3{\mathrm{<figure-callout\; id="1"\; label="c"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">c</figure-callout>}}_1\mathrm{\&Delta;d}}{e}^{-{\mathrm{\&epsiv;}}_4{c}_2\mathrm{\&Delta;d}})---\left(6\right)$

I wherein ₀Transmitted light intensity when only organized background absorbs, Δ d is the change of the transmission distance brought to blood engorgement by depletion of blood.At an easy rate, definition bias light density D _B: D _B=ln (I ₀/ I _B); Thereby, can obtain:

D(660)-D _B(660)＝ε ₁C ₁Δd+ε ₂C ₂Δd(7)

D(805)-D _B(805)＝ε ₃C ₁Δd+ε ₄C ₂Δd(8)

ε wherein ₃=ε ₄=ε, therefore (4) formula develops and is:

SaO ₂＝A×(D(660)-D _B(660))/(D(805)-D _B(805))+B(9)

A, B are the same.Formula (9) is the fundamental formular that detects blood oxygen saturation.

In common detection,, be the comparison difficulty often because will obtain the exact value of this isobestic point without the infrared light of wavelength 805nm, the error of bringing thus is also bigger; Infrared light wavelength commonly used is about 940nm, near HbO this wavelength ₂, Hb absorption variations all compare gently, error is often smaller.Owing to adopted the infrared light of wavelength 940nm, the ε in the formula (8) ₃≠ ε ₄, formula (10) further develops and is:

Spo ₂＝(A×R+B)/(C×R+D)(10)

Wherein, SpO ₂Be detecting sphygmus and blood oxygen saturation, it is approximately SaO ₂

A=ε ₁B=-ε ₂C=ε ₄-ε ₃D=ε ₁-ε ₂And

$R=\frac{D\left(660\right)-D_B\left(660\right)}{D\left(940\right)-D_B\left(940\right)},---\left(11\right)$

By formula as can be known, R and blood oxygen saturation are one to one.And D=LnI ₀/ I=ε cd

Then $R=\frac{\ln I_{R0}/I_{\mathrm{RM}}-\ln I_{R0}/I_{\mathrm{Rm}}}{\ln I_{I0}/I_{\mathrm{LM}}-\ln I_{I0}/I_{\mathrm{Im}}}=\frac{\ln I_{\mathrm{Rm}}/I_{\mathrm{RM}}}{{\ln I}_{\mathrm{Im}}/I_{\mathrm{LM}}}---\left(12\right)$

In the formula, I _RMBe HONGGUANG max transmissive light, I _RmBe the minimum transillumination of HONGGUANG, I _R0Be the HONGGUANG incident illumination.I _IMBe infrared max transmissive light, I _ImBe infrared minimum transillumination, I _I0Be infrared incident illumination.For HONGGUANG,

$\ln I_{\mathrm{Rm}}/I_{\mathrm{RM}}=\ln (1-\frac{I_{\mathrm{RM}}-I_{\mathrm{Rm}}}{I_{\mathrm{RM}}}),---\left(13\right)$

When pulsating quantity/DC quantity is (I _RM-I _Rm)/I _RMHour,

So R can be write as following form:

$R=\frac{{\mathrm{Red}}_{\mathrm{AC}}/{\mathrm{Red}}_{\mathrm{DC}}}{{\mathrm{Ir}}_{\mathrm{AC}}/{\mathrm{Ir}}_{\mathrm{DC}}}$

If therefore know that the waveform of two kinds of transilluminations in a complete pulse wave just can calculate the R value.

1, motion noise is to the influence of measurement result

Can draw from top formula, after luminous tube was fixing, having only R in (10) formula was variable, and the DC component of HONGGUANG and infrared light also is more stable in a period of time, and influenced the AC compounent of the principal element of R for this two-way light.The method of traditional calculating AC compounent is exactly to obtain by the maximum of seeking two-way light and the method for minima.But there is very big defective in this method, is in as patient under the condition of motion.The searching of waveform is very difficult to this moment, owing to error can occur to the searching of pulse wave crest, trough, therefore the ratio of the alternating current-direct current that obtains also may be wrong, and the measurement result of blood oxygen and pulse frequency is difficult to guarantee correctly.As shown in Figure 4, situations such as baseline drift, pulse noise, step noise can take place in the pulse wave under the moving condition, are difficult to judge exactly the value of two-way light this moment.

2, how to abate the noise to the influence of measurement result

For addressing this problem, the present invention has designed a kind of algorithm that signal is carried out differential and asymptotic intergration, can be contained under the moving condition noise effectively to the interference of signal waveform, and can prove blood oxygen levels that this algorithm obtains theoretically and be of equal value that therefore algorithm of the present invention has solved the inaccurate problem of measurement moving condition under not having to adopt the algorithm of seeking the waveform AC compounent under the situation of motion artifacts.Introduce the step and the principle of this algoritic module below by a specific embodiment.

The two-way light data that measure can obtain the direct current ratio of two-way light through after the normalization

Blood oxygen waveform after the normalization can be regarded waveform and noise synthetic ideally as, and blood oxygen waveform ideally, and no matter red waveforms still be the infrared light waveform, can regard the synthesizing of cosine wave of different frequency range as

Red＝a ₀cos(ωt)+a ₁cos(2ωt)+Λ+a _n-1cos(ωt)+n _Red＝Red _sig+n _Red(14)

Ir＝b ₀cos(ωt)+b ₁cos(2ωt)+Λ+b _n-1cos(nωt)+n _Ir＝Ir _sig+n _Ir(15)

A wherein ₀, a ₁... a _N-1Be HONGGUANG frequency spectrum n order component, n _RedNoise composition in the HONGGUANG has wherein comprised white noise and nonwhite noise, b ₀, b ₁... b _N-1Be HONGGUANG frequency spectrum n order component, n _IrNoise composition in the infrared light has wherein comprised white noise and nonwhite noise.Present embodiment adopts cosine wave as the wave spectrum expanded function, in fact can also use sine wave or other can constitute the function of complete orthogonal system as the wave spectrum expanded function.

Brief description is handled the differential albefaction of noise composition in the pulse wave.

Noise composition wherein may be caused by multiple situation, comprises because kinetic baseline drift noise is shaken the noise that step changes that is similar to that causes suddenly, and the noise of high frequency.

2.1 differential is removed baseline drift

To (14) (15) difference differential,

$\frac{\mathrm{dRed}}{\mathrm{dt}}=\frac{{\mathrm{dRed}}_{\mathrm{sig}}}{\mathrm{dt}}+\frac{{\mathrm{dn}}_{\mathrm{Red}}}{\mathrm{dt}}---\left(16\right)$

$\frac{\mathrm{dIr}}{\mathrm{dt}}=\frac{{\mathrm{dIr}}_{\mathrm{sig}}}{\mathrm{dt}}+\frac{{\mathrm{dn}}_{\mathrm{Ir}}}{\mathrm{dt}}---\left(17\right)$

DC component in HONGGUANG or the infrared light is zero behind the differential.The pulse wave that comprises baseline drift as shown in Figure 5 because the calculating of blood oxygen all is to take the employing data in a period of time to carry out, in the slowly baseline drift of this section in the time, is converted into constant through behind the differential, and then noise can be expressed as,

Therefore,, can remove, promptly remove kinetic baseline drift noise, as shown in Figure 6 by normalization for the constant component behind the noise differential.

2.2 differential is eliminated the step noise

For the unexpected extruding of shaking the medium that may cause, the step noise that causes as shown in Figure 7, then to behind its differential, becomes the impulse function of similar delta-function, as shown in Figure 8, then can take the filtering of 3 intermediate values to this, eliminates the step noise.

2.3 differential is eliminated pulse noise

The sampled value sudden change that causes for shake promptly is similar to the pulse noise of delta-function, shown in figure (9), after the digitized differential, becomes the function of positive and negative dipulse, shown in figure (10), can adopt the in addition filtering of 5 intermediate values to this, eliminates pulse noise.

From the above mentioned, after differential and medium filtering processing, can eliminate baseline drift, step noise and pulse noise in HONGGUANG and the infrared light pulse wave signal to pulse wave.

2.4 integration is replied pulse waveform

The pulse wave that comprises noise shown in Figure 4 is through after top differential and the median function processing, as shown in figure 11, (18), (19) formula are carried out integration again, reply pulse waveform, the waveform of this moment has become smooth, aforementioned differential to waveform is handled and is equivalent to pulse wave has been carried out the albefaction processing, has eliminated the nonwhite noise in the noise.

$\underset{t_0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">t</figure-callout>}}_1}{\&Integral;}}\frac{\mathrm{dRed}}{\mathrm{dt}}=\underset{t_0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">t</figure-callout>}}_1}{\&Integral;}}\frac{\mathrm{dRe}{d}_{\mathrm{sig}}}{\mathrm{dt}}+\underset{t_0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">t</figure-callout>}}_1}{\&Integral;}}\frac{{\mathrm{dn}}_{\mathrm{Red}}}{\mathrm{dt}}$ ,

Red＝Red _sig+n _Red‘’(20)

$\underset{t_0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">t</figure-callout>}}_1}{\&Integral;}}\frac{\mathrm{dIr}}{\mathrm{dt}}=\underset{t_0}{\overset{t_1}{\&Integral;}}\frac{d{\mathrm{Ir}}_{\mathrm{sig}}}{\mathrm{dt}}+\underset{t_0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="t"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">t</figure-callout>}}_1}{\&Integral;}}\frac{{\mathrm{dn}}_{\mathrm{Ir}}}{\mathrm{dt}}$ ,

Ir＝Ir _sig+n _Ir‘’(21)

N wherein _Red' ' and n _Ir' ' handles the remaining white noise part in back for the pulse wave noise through albefaction.

2.5. area integral recursive algorithm principle

Respectively to (20), (21) formula integration, and obtain ratio

If the noise in a period of time can be regarded as white noise, then its integration is zero, and following formula is:

Thus, need only the long enough time of integration, the integration of noise levels off to zero, can replace from waveform seeking the two-way light interchange data that value obtains this moment by the integration data of a period of time, and because this method has been eliminated interference of noise, thereby under moving condition, can obtain good measurement effect.Figure 12 is the integration figure of pulse wave.

More than be that blood oxygen saturation with tested object in a period of time does not change to the hypothesis prerequisite.Long more then measurement effect is good more the time of integration so, the approaching more and truth of result.And when the blood oxygen saturation of interior tested object changes, the time of integration is long not to have benefit on the contrary, causing measuring sensitivity descends, consider serious situation the most, if integration has just comprised all measurement data from measuring beginning, then the data saturated phenomenon can take place, measurement data new after a period of time is very little to result's influence, thus the function of the real-time measurement that weakens.

For addressing the above problem, integration only carries out (the integration period was generally 2 to 3 seconds) in a period of time, for keeping the effect of measuring early stage and don't making it excessive, introduces forgetting factor λ simultaneously.

$\frac{{\mathrm{Red}}_{\mathrm{AC}}}{{\mathrm{Ir}}_{\mathrm{AC}}}=\frac{{\mathrm{Red}}_{{\mathrm{AC}}_0}+\mathrm{\&lambda;}{\mathrm{<figure-callout\; id="1"\; label="Red\; AC"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">Red</figure-callout>}}_{{\mathrm{AC}}_{}}+\mathrm{\&Lambda;}+{\mathrm{\&lambda;}}^n{\mathrm{Red}}_{{\mathrm{AC}}_n}}{{\mathrm{Ir}}_{{\mathrm{AC}}_0}+\mathrm{\&lambda;}{\mathrm{<figure-callout\; id="1"\; label="Ir\; AC"\; filenames="C20051012126900141.png,C20051012126900152.png"\; state="\{\{state\}\}">Ir</figure-callout>}}_{{\mathrm{AC}}_{}}+\mathrm{\&Lambda;}+{\mathrm{\&lambda;}}^n{\mathrm{Ir}}_{{\mathrm{AC}}_n}}---\left(24\right)$

When 0＜λ＜1, through the several times iteration, the data of front are not influence, and rule of thumb λ gets 0.8 relatively rationally.

System's realization flow:

After system powers on, carry out hardware initialization earlier, cpu system self check, and program initialization; Enter kernel control module after finishing above process;

The real-time image data of probe assembly, the data storage that measures are carried out the basis that date processing calculates blood oxygen, pulse frequency parameter to the data buffer zone as core algorithm;

Logic control module carries out different control to hardware according to measured value in each state, and control AD sampling.Comprise inner AD and outside AD, i.e. analogue signal input sample module among the figure.And to the control of luminous tube drive current, to the control of biasing circuit and gain;

The motion noise cancellation module is a nucleus module, and this module is eliminated nonwhite noise by the differential of pulse wave is handled respectively, calculates pulse frequency simultaneously, calculates blood oxygen by the integration recursive algorithm again.

Claims (9)

1, a kind of Oximetry data processing method of eliminating motion artifacts comprises the steps:

A. after probe will be detected has optical signal that pulse involves blood oxygen saturation information and is converted to the signal of telecommunication, this signal of telecommunication carried out processing and amplifying after, be admitted to A/D and change and be transformed into digital signal;

The b.A/D modular converter is the analog signal conversion through processing and amplifying a digital signal;

The digital signal of c.A/D modular converter output is admitted to single-chip microcomputer and carries out date processing and calculating;

It is characterized in that: date processing and calculating that described single-chip microcomputer carries out comprise the steps:

D. pulse wave and blood oxygen waveform are regarded as the stack of certain waveform and noise, thus the HONGGUANG of reflection blood oxygen waveform characteristic and the time series of infrared light pulse wave signal are expressed as the synthetic of different frequency range wave spectrum, promptly the time series of initial data being carried out wave spectrum launches, in the wave spectrum space, the motion noise of blood oxygen waveform is for conversion into a constant;

E. the wave spectrum expression formula of above-mentioned initial data being differentiated, is zero behind the motion noise differential as constant, has promptly filtered motion noise by differentiating;

F. the slow baseline drift of initial data shows as an oblique line in wave amplitude time coordinate system, has the waveform of physical significance to be superimposed upon on this oblique line, filters the baseline drift noise by normalized;

G. the noise that shake is caused shows as the pulse noise of step noise or delta-function in wave amplitude time coordinate system, the pulse noise of this step noise and delta-function is got intermediate value and adjacent 5 data points by adjacent 3 data points respectively and got intermediate value and be filtered;

H. by the HONGGUANG of having eliminated baseline drift, step noise and pulse noise in the wave spectrum space and infrared light pulse wave signal being carried out described pulse wave and blood oxygen waveform are reduced in the area integral computing of sampling period;

I. the forgetting factor λ that introduces value between 0 and 1 eliminates the data saturated phenomenon of long-time integration, promptly to measure constantly and the result of calculation of previous reasonable period carry out the result of calculation that iteration obtains this measurements moment, the approaching more described data constantly of measuring are big more to the contribution of result of calculation during iteration;

J. the pulse wave of step I acquisition and blood oxygen saturation calculating result outwards export by the serial ports that is electrically connected with described single-chip microcomputer.

2, the Oximetry data processing method of elimination motion artifacts according to claim 1 is characterized in that: the described unfolded function of wave spectrum that is used for is other functions that sinusoidal wave function, cosine wave function maybe can constitute complete orthogonal system.

3, the Oximetry data processing method of elimination motion artifacts according to claim 1 is characterized in that: the integration period of described step h is 2 to 3 seconds; Described forgetting factor λ value is 0.8, and the iterative relation formula of introducing this forgetting factor λ is as follows:

$\frac{{\mathrm{Red}}_{\mathrm{AC}}}{{\mathrm{Ir}}_{\mathrm{AC}}}=\frac{{\mathrm{Red}}_{A{C}_0}+\mathrm{\&lambda;}{\mathrm{Red}}_{A{C}_1}+\&CenterDot;\&CenterDot;\&CenterDot;+{\mathrm{\&lambda;}}^n{\mathrm{Red}}_{{\mathrm{AC}}_n}}{{\mathrm{Ir}}_{{\mathrm{AC}}_0}+\mathrm{\&lambda;}{I{r}_{\mathrm{AC}}}_1+\&CenterDot;\&CenterDot;\&CenterDot;+{\mathrm{\&lambda;}}^n{\mathrm{Ir}}_{{\mathrm{AC}}_n}}.$

4, the Oximetry data processing method of elimination motion artifacts according to claim 1 is characterized in that: described probe utilizes the two-way light transmission to organize end to carry out sampled measurements; Wherein one the tunnel is HONGGUANG, and corresponding data integrate result is that the red light intensity that receives in this section period exchanges peak-to-peak value, and another road is an infrared light, and corresponding data integrate result is that the infrared light intensity that receives in this section period exchanges peak-to-peak value.

5, the Oximetry data processing method of elimination motion artifacts according to claim 1 is characterized in that: comprised also before step a that system powers on, the sampling set-up procedure of hardware initialization, cpu system self check and program initialization.

6, a kind of blood oxygen measuring device that can eliminate motion artifacts, comprise: the probe assembly that is linked in sequence, signal processing and amplifying assembly, the A/D modular converter, the single-chip data processing components, Logic control module and power circuit, it is characterized in that: described single-chip data processing components comprises the motion artifacts cancellation module, this motion artifacts cancellation module comprises the pulse wave parameter calculating module of the unidirectional connection of order again, the differential processing module, the medium filtering module, integration recovery module and area integral recurrence module, the output result of described area integral recurrence module is used to calculate pulse wave and blood oxygen saturation, and the outfan of described motion artifacts cancellation module then connects A/D modular converter and logic control and functional module respectively.

7, the blood oxygen measuring device that can eliminate motion artifacts according to claim 6, it is characterized in that: described probe assembly comprises HONGGUANG, infrared light supply and light drive circuit thereof and light-receiving tube.

8, the blood oxygen measuring device that can eliminate motion artifacts according to claim 6 is characterized in that: described signal enlargement processing section branch comprises differential amplifier circuit and signal processing circuit.

9, the blood oxygen measuring device that can eliminate motion artifacts according to claim 6, it is characterized in that: described logic control and functional module comprise security function and communication function; Described security function comprises by pick off and comes off, searches for pulse, the oversize and blood oxygen saturation of search pulse descend measurement Status Flag of being formed and the module self check of finishing error report; Affiliated communication function comprises receiving orders and sends data.

Images