CN101972148A - Disturbance elimination method of near infrared brain function detection based on empirical mode decomposition - Google Patents

Abstract

Translated fromChinese

基于经验模态分解的近红外脑功能检测的扰动消除方法，属于光学领域，本发明为解决采用低通滤波无法全面有效去除脑功能检测时生理扰动；采用自适应滤波技术存在需要借助额外的设备、结构复杂的问题。本发明方法包括以下步骤：一、在待测脑组织的头皮表面放置由双波长光源和检测器构成的近红外探头，获得光密度变化量时间序列：

和

；二、采用修正朗伯比尔定律获取氧合血红蛋白浓度变化量时间序列Δ[HbO₂](t)和还原血红蛋白浓度变化量时间序列Δ[HHb](t)；三、对Δ[HbO₂](t)和Δ[HHb](t)分别进行经验模态分解，获得所有的IMF分量；四、对所有IMF分量进行希尔伯特变换，将瞬时频率处于正常人呼吸频率和心脏跳动频率范围内的IMF分量剔除，以消除近红外脑功能检测时的生理扰动。

The disturbance elimination method of near-infrared brain function detection based on empirical mode decomposition belongs to the field of optics. The present invention solves the problem that low-pass filtering cannot fully and effectively remove physiological disturbances during brain function detection; the use of adaptive filtering technology requires the use of additional equipment , problems with complex structures. The method of the present invention comprises the following steps: 1. Place a near-infrared probe consisting of a dual-wavelength light source and a detector on the scalp surface of the brain tissue to be measured to obtain a time series of optical density changes:

and

; 2. Obtain the time series Δ[HbO₂ ](t) of oxygenated hemoglobin concentration change and the time series Δ[HHb](t) of reduced hemoglobin concentration change by using amended Lambert-Beer's law; 3. For Δ[HbO₂ ] (t) and Δ[HHb](t) are subjected to empirical mode decomposition to obtain all IMF components; 4. Hilbert transform is performed on all IMF components, and the instantaneous frequency is in the range of normal human breathing frequency and heart beating frequency The internal IMF component is eliminated to eliminate the physiological disturbance in near-infrared brain function detection.

Description

The disturbance removing method of the Near-infrared Brain Function detection of decomposing based on empirical modal

Technical field

The present invention relates to disturbance removing method, belong to optical field based on the Near-infrared Brain Function detection of empirical modal decomposition.

Background technology

Near-infrared spectrum technique (NIRS) can provide the information of the cerebral cortex blood oxygen metabolism in the cerebration process---HbO2 Oxyhemoglobin concentration change (Δ [HbO₂]) and reduced hemoglobin concentration change (Δ [HHb]), can be used for the detection of cerebration.With other the brain function detection method as: functional magnetic resonance resonance, magnetoencephalography, positron emission tomography and EECG are compared, easy to use, easy enforcement that near-infrared spectrum technique has, temporal resolution height, safety, advantage such as cheap.Yet, utilize near-infrared spectrum technique to bring out the detection of when excitation cerebration, can be subjected to the influence that the physiological activity such as the heart of human body are beated, breathed, be referred to as the physiology disturbance.This physiology disturbance not only appears in the outer cerebral tissue such as scalp, skull and cerebrospinal fluid, also appears in the deep layer cerebral tissue such as ectocinerea and alba, has had a strong impact on the accurate measurement of cerebration.

At the influence of physiology disturbance to the Near-infrared Brain Function detection, the most direct processing method is to utilize low-pass filtering technique.Low-pass filtering technique can filter the disturbance that heartbeat causes effectively, because the forcing frequency that heartbeat causes is apparently higher than the cerebration signal.But low-pass filtering technique can't effectively filter the disturbance that breathing causes, this is because the forcing frequency that breathing causes is very low, and low excessively cut-off frequency has also caused the distortion of cerebration signal in this type of turbulent while of elimination.Auto-adaptive filtering technique is as one of turbulent removing method of physiology, showed de-noising characteristic preferably, can reduce the influence that the physiology disturbance detects at brain function near-infrared spectrum technique, but adaptive technique need be by means of pulse blood oxygen instrument or extra path channels, complex structure.

Summary of the invention

The present invention seeks in order to solve the multiple physiology disturbance when adopting low-pass filtering can't remove brain function comprehensively and effectively to detect; The turbulent method of physiology when adopting auto-adaptive filtering technique to eliminate brain function to detect exists need be by extra equipment, baroque problem, and a kind of disturbance removing method of Near-infrared Brain Function detection of decomposing based on empirical modal is provided.

The inventive method may further comprise the steps:

Step 1, place the near-infrared probe that constitutes by double-wavelength light source and detector in the scalp surface of cerebral tissue to be measured, diffuse-reflectance light intensity under the detector recording brain rest state and brain are in the diffuse-reflectance light intensity of bringing out when excitation, the time series of the optical density variable quantity when obtaining two different wave lengths:

With

T is the time, t=1, and 2 ..., N;

The time series ofstep 2, the optical density variable quantity that obtains according tostep 1, and adopt and revise the time series Δ [HbO that langbobier law obtains HbO2 Oxyhemoglobin concentration change amount₂] (t) and the time series Δ [HHb] of reduced hemoglobin concentration change amount (t);

$\mathrm{\Δ}\left[{\mathrm{HbO}}_2\right]\left(t\right)=\frac{({\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_1\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_2}\left(t\right)/\mathrm{DPF})-({\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_2\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_1}\left(t\right)/\mathrm{DPF})}{r({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_2\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_1\right)-{\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_1\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_2\right))},$

$\mathrm{\Δ}\left[\mathrm{HHb}\right]\left(t\right)=\frac{({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_2\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_1}\left(t\right)/\mathrm{DPF})-({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_1\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_2}\left(t\right)/\mathrm{DPF})}{r({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_2\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_1\right)-{\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_1\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_2\right))},$

Wherein, ε_HHb(λ₁) for the wavelength of probe light source be λ₁The time extinction coefficient,

For the wavelength of probe light source is λ₂The time extinction coefficient,

R is the air line distance of light source to detector,

DPF is the differential path factor,

Time series Δ [the HbO of step 3, HbO2 Oxyhemoglobin concentration change amount thatstep 2 is obtained₂] (t) and the time series Δ [HHb] of reduced hemoglobin concentration change amount (t) carry out empirical modal respectively and decompose, obtain all IMF (intrinsicmode function, in accumulate mode function) component;

Step 4, all IMF components that step 3 is obtained carry out Hilbert transform, ask for the instantaneous frequency of each IMF component, instantaneous frequency is in IMF component rejection in normal person's respiratory frequency and the heartbeat frequency range, the physiology disturbance when eliminating the Near-infrared Brain Function detection.

Advantage of the present invention: the inventive method only need utilize easy probe can realize effectively eliminating the physiology disturbance, the inventive method adopts empirical modal to decompose this time frequency analysis method, handle the non-stationary nonlinear properties, complicated original signal is resolved into limited simple component, be called the IMF component, it has good Hilbert transform characteristic, makes instantaneous frequency to calculate.This decomposition method makes instantaneous frequency have the actual physical meaning, thereby can determine effectively rejecting of physiology disturbance in conjunction with human body physiological parameter.Δ [HbO when empirical modal decomposition and Hilbert transform can be applicable to cerebration₂] and the reconstruct of Δ [HHb], can eliminate more than 90% by the physiology disturbance of breathing and heartbeat causes.

Description of drawings

To be the present invention carry out the flow chart that decomposes based on empirical modal to the IMF component to Fig. 1.

The specific embodiment

The specific embodiment one: below in conjunction with Fig. 1 present embodiment is described, the present embodiment method may further comprise the steps:

Step 1, place the near-infrared probe that constitutes by double-wavelength light source and detector in the scalp surface of cerebral tissue to be measured, diffuse-reflectance light intensity under the detector recording brain rest state and brain are in the diffuse-reflectance light intensity of bringing out when excitation, the time series of the optical density variable quantity when obtaining two different wave lengths:

With

T is the time, t=1, and 2 ..., N;

The time series ofstep 2, the optical density variable quantity that obtains according tostep 1, and adopt and revise the time series Δ [HbO that langbobier law obtains HbO2 Oxyhemoglobin concentration change amount₂] (t) and the time series Δ [HHb] of reduced hemoglobin concentration change amount (t);

$\mathrm{\Δ}\left[{\mathrm{HbO}}_2\right]\left(t\right)=\frac{({\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_1\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_2}\left(t\right)/\mathrm{DPF})-({\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_2\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_1}\left(t\right)/\mathrm{DPF})}{r({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_2\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_1\right)-{\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_1\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_2\right))},$

$\mathrm{\Δ}\left[\mathrm{HHb}\right]\left(t\right)=\frac{({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_2\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_1}\left(t\right)/\mathrm{DPF})-({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_1\right)\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_2}\left(t\right)/\mathrm{DPF})}{r({\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_2\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_1\right)-{\mathrm{\ϵ}}_{{\mathrm{HbO}}_2}\left({\mathrm{\λ}}_1\right){\mathrm{\ϵ}}_{\mathrm{HHb}}\left({\mathrm{\λ}}_2\right))},$

Wherein, ε_HHb(λ₁) for the wavelength of probe light source be λ₁The time extinction coefficient, depend on wavelength and specific absorbing material, be constant,

For the wavelength of probe light source is λ₂The time extinction coefficient,

R is the air line distance of light source to detector,

DPF is the differential path factor, DPF (differential pathlength factor,), can measure with frequency domain spectroscopy or time-resolved spectroscopy method, also can obtain by Monte Carlo simulation calculation, little with wavelength change, be constant, adult brain tissue value 5.6, child's value 4.3.

Time series Δ [the HbO of step 3, HbO2 Oxyhemoglobin concentration change amount thatstep 2 is obtained₂] (t) and the time series Δ [HHb] of reduced hemoglobin concentration change amount (t) carry out empirical modal respectively and decompose, obtain all IMF components;

Step 4, all IMF components that step 3 is obtained carry out Hilbert transform, ask for the instantaneous frequency of each IMF component, instantaneous frequency is in IMF component rejection in normal person's respiratory frequency and the heartbeat frequency range, the physiology disturbance when eliminating the Near-infrared Brain Function detection.

The technical scheme of present embodiment is achieved in that the sonde configuration that utilizes single light source list detector, and light source adopts double-wavelength light source (λ₁=750nm, λ₂=830nm), light source is 45mm to the air line distance (light source detection device spacing) of detector.Light source detection device spacing is approximately the twice of near infrared light investigation depth, and investigation depth can reach 20-22mm when light source detection device spacing was 45mm, and setting can make near infrared light can effectively penetrate cerebral cortex like this.Change the optical density variation that obtains into HbO2 Oxyhemoglobin Δ [HbO by revising langbobier law₂] and the concentration change amount Δ [HHb] of reduced hemoglobin, the Δ [HbO of this moment₂] and Δ [HHb] doping physiology turbulent noise even flooded by turbulent noise.Utilize EMD (empirical modal decomposition) with Δ [HbO₂] be decomposed into IMF component with Δ [HHb] with different local features, and the IMF component is carried out Hilbert transform ask for instantaneous frequency.Utilize the instantaneous frequency of all IMF components, the IMF component that can determine physiology disturbance correspondence in conjunction with normal person's breathing and heartbeat frequency.Reject the IMF component of physiology disturbance correspondence, reconstruct cerebration signal.

Wherein, two kinds of wavelength sending of the described double-wavelength light source ofstep 1 are respectively λ₁=750nm, λ₂=830nm.

The time series of optical density variable quantity in thestep 1

Obtain by following formula:

$\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_1}\left(t\right)=\log I_{\mathrm{base}}\left({\mathrm{\λ}}_1\right)/I_{\mathrm{stim}}\left({\mathrm{\λ}}_1\right),$

Wherein: I_Base(λ₁) for the wavelength of probe light source be λ₁The time, brain is in the rest state output intensity in following time, and in the initial moment, brain is under the rest state, and the diffuse-reflectance light intensity that the record detector receives is as test benchmark.

I_Stim(λ₁) for the wavelength of probe light source be λ₁The time, brain is in the output intensity when bringing out excitation,

At near infrared band HbO₂With HHb be main absorber, and there is significant difference in its absorption spectra.Therefore, based on the Near-infrared Brain Function detection of continuous spectrum technology, mainly be to measure HbO2 Oxyhemoglobin (HbO₂) and the concentration change of reduced hemoglobin (HHb).

Detect for brain function, adopt dual wavelength continuous light near-infrared measuring system, the time series of optical density variable quantity

Obtain by following formula:

$\mathrm{\Δ}{\mathrm{OD}}_{{\mathrm{\λ}}_2}\left(t\right)=\log I_{\mathrm{base}}\left({\mathrm{\λ}}_2\right)/I_{\mathrm{stim}}\left({\mathrm{\λ}}_2\right),$

Wherein: I_Base(λ₂) for the wavelength of probe light source be λ₂The time, brain is in the rest state output intensity in following time,

I_Stim(λ₂) for the wavelength of probe light source be λ₂The time, brain is in the output intensity when bringing out excitation.

In the step 3 to the time series Δ [HbO of HbO2 Oxyhemoglobin concentration change amount₂] (t) (t) to carry out the process that empirical modal decomposes identical with the time series Δ [HHb] of reduced hemoglobin concentration change amount, it is a kind of analytical method of non-linear, nonstationary time series that empirical modal decomposes, it can carry out linearisation to original series, tranquilization is handled, and keeps the feature of original series self in catabolic process.The time scale feature of its basis signal itself, with signal decomposition is to contain the different time yardstick and satisfy in a group of following two definite conditions and accumulate mode function (IMF): 1. in whole data sequence, the number of signal extreme point must differ one identical or at most with the number of signal zero crossing; 2. at any time on, the average of signal local maximum envelope and signal local minimum envelope is zero.

Described optical density signal is converted to Δ [HbO₂] (t) and Δ [HHb] (t) carry out EMD then and decompose, the effectiveness of EMD method is to decompose Δ [HbO₂] (t) and Δ [HHb] (t) rather than directly decompose and comprise the turbulent signal that diffuses of physiology because diffuse signal and Δ [HbO₂] (t) and Δ [HHb] be index relation (t), the EMD algorithm decomposes the signal that diffuses can't realize separating fully the IMF component of physiology disturbance correspondence, thereby can't eliminate the physiology disturbance.

Below with Δ [HbO₂] (t) and Δ [HHb] (t) be referred to as C_Ij(t), which IMF component i is, i=1, and 2 ..., n, j is the estimation number of times, initialization i=1, j=1 is to time series C_Ij(t) carry out empirical modal and decompose the acquisition process that obtains all IMF components:

Step 1, employing local extremum method are determined the hunting time sequence C_Ij(t) all maximum and minimum make up time series C with cubic spline interpolation respectively to all maximum value minimums that obtain_Ij(t) coenvelope line e_Max(t) and lower envelope line e_Min(t);

Step 2, obtain the average of upper and lower envelope

$\stackrel{\‾}{e\left(t\right)}=\frac{e_{\max }\left(t\right)+e_{\min }\left(t\right)}{2},$

Estimate h the j time of step 3, i IMF component of acquisition time sequence_Ij(t):

$h_{\mathrm{ij}}\left(t\right)=C_{\mathrm{ij}}\left(t\right)-\stackrel{\‾}{e\left(t\right)},$

Step 4, judge whether following formula is set up:

ε＞0 wherein, and fully near 0,

Judged result is for being, execution in step 5,

Judged result makes j=j+1, C for not_{I (j+1)}(t)=h_Ij(t), and return execution instep 1,

Step 5, obtain i IMF component: C_i(t)=h_IjAnd obtain i residual error (t):

r_i(t)＝C_ij(t)-h_ij(t)，

Step 6, i residual error r of judgement_i(t) whether be monotonic function,

Judged result makes i=i+1 for not, j=1, and return execution instep 1,

Judged result is for being to finish time series C_Ij(t) carry out the empirical modal decomposition and obtain all IMF components: C_i(t).

The acquisition methods of the instantaneous frequency of IMF component is in the step 4:

Step 41, acquisition IMF component C_i(t) Hilbert transform y (t):

$y\left(t\right)=H\left[C_i\left(t\right)\right]=\frac{1}{\mathrm{\π}}P{\∫}_{-\∞}^{\∞}\frac{C_i\left(t^{\′}\right)}{t-t^{\′}}d{t}^{\′},$

Wherein, P represents Cauchy's principal value,

Step 42, IMF component C_i(t) analytic signal is z (t)=C (t)+iy (t)=a (t) exp[i θ (t)], wherein, a (t) is an instantaneous amplitude, θ (t) is a phase function,

Step 43, obtain IMF component C_i(t) instantaneous frequency f (t) is:

Normal person's respiratory frequency scope is 0.15Hz～0.4Hz, and the heartbeat frequency range is 1.0Hz～1.7Hz.

The instantaneous frequency of IMF component has embodied the internal feature of IMF component, according to normal person's respiratory frequency and heartbeat frequency, determines the IMF component of physiology disturbance correspondence.With the IMF component rejection of physiology disturbance correspondence, utilize other IMF component reconstruct to obtain the turbulent cerebration signal of physiology.

Described optical density signal is converted to Δ [HbO₂] (t) and Δ [HHb] (t) carry out EMD then and decompose, the effectiveness of EMD method is to decompose Δ [HbO₂] (t) and Δ [HHb] (t) rather than directly decompose and comprise the turbulent signal that diffuses of physiology because diffuse signal and Δ [HbO₂] (t) and Δ [HHb] be index relation (t), the EMD algorithm decomposes the signal that diffuses can't realize separating fully the IMF component of physiology disturbance correspondence, thereby can't eliminate the physiology disturbance.

Claims (6)

Translated fromChinese

1.基于经验模态分解的近红外脑功能检测的扰动消除方法，其特征在于，它包括以下步骤：1. the disturbance elimination method based on the near-infrared brain function detection of empirical mode decomposition, it is characterized in that, it comprises the following steps:步骤一、在待测脑组织的头皮表面放置由双波长光源和检测器构成的近红外探头，检测器记录大脑安静状态下的漫反射光强和大脑处于诱发激励时的漫反射光强，以获得两个不同波长时的光密度变化量的时间序列：和

t为时间，t＝1，2，...，N；Step 1. Place a near-infrared probe consisting of a dual-wavelength light source and a detector on the surface of the scalp of the brain tissue to be tested. The detector records the diffuse reflection light intensity of the brain in a quiet state and the diffuse reflection light intensity of the brain when it is in the evoked excitation. Obtain a time series of optical density changes at two different wavelengths: and

t is time, t=1, 2,..., N;步骤二、根据步骤一获得的光密度变化量的时间序列、并采用修正朗伯比尔定律获取氧合血红蛋白浓度变化量的时间序列Δ[HbO₂](t)和还原血红蛋白浓度变化量的时间序列Δ[HHb](t)；Step 2. According to the time series of optical density changes obtained in step 1, the time series Δ[HbO₂ ](t) of oxygenated hemoglobin concentration changes and the time series of reduced hemoglobin concentration changes are obtained by using the amended Lambert-Beer's law Δ[HHb](t);$\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}<span><span>[</span>[</span>{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>]</span>]</span><span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\frac{<span><span>(</span>(</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{\mathrm{<span><span>HHb</span>Hb</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span>\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}{\mathrm{<span><span>OD</span>OD</span>}}_{{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>/</span>/</span>\mathrm{<span><span>DPF</span>DPF</span>}<span><span>)</span>)</span><span><span>-</span>-</span><span><span>(</span>(</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{\mathrm{<span><span>HHb</span>Hb</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span>\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}{\mathrm{<span><span>OD</span>OD</span>}}_{{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>/</span>/</span>\mathrm{<span><span>DPF</span>DPF</span>}<span><span>)</span>)</span>}{\mathrm{<span><span>r</span>r</span>}<span><span>(</span>(</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{\mathrm{<span><span>HHb</span>Hb</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span><span><span>-</span>-</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{\mathrm{<span><span>HHb</span>Hb</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span><span><span>)</span>)</span>}<span><span>,</span>,</span>$$\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}<span><span>[</span>[</span>\mathrm{<span><span>HHb</span>Hb</span>}<span><span>]</span>]</span><span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\frac{<span><span>(</span>(</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span>\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}{\mathrm{<span><span>OD</span>OD</span>}}_{{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>/</span>/</span>\mathrm{<span><span>DPF</span>DPF</span>}<span><span>)</span>)</span><span><span>-</span>-</span><span><span>(</span>(</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span>\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}{\mathrm{<span><span>OD</span>OD</span>}}_{{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>/</span>/</span>\mathrm{<span><span>DPF</span>DPF</span>}<span><span>)</span>)</span>}{\mathrm{<span><span>r</span>r</span>}<span><span>(</span>(</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{\mathrm{<span><span>HHb</span>Hb</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span><span><span>-</span>-</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{{\mathrm{<span><span>HbO</span>HbO</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span>{\mathrm{<span><span>\&epsiv;</span>\&epsiv;</span>}}_{\mathrm{<span><span>HHb</span>Hb</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span><span><span>)</span>)</span>}<span><span>,</span>,</span>$其中，ε_HHb(λ₁)为探头光源的波长为λ₁时的消光系数，Among them, ε_HHb (λ₁ ) is the extinction coefficient when the wavelength of the probe light source is λ₁ ,

为探头光源的波长为λ₂时的消光系数，

is the extinction coefficient when the wavelength of the probe light source is_λ2 ,r为光源到检测器的直线距离，r is the linear distance from the light source to the detector,DPF为差分路径因子，DPF is the differential path factor,步骤三、对步骤二获得的氧合血红蛋白浓度变化量的时间序列Δ[HbO₂](t)和还原血红蛋白浓度变化量的时间序列Δ[HHb](t)分别进行经验模态分解，获得所有的IMF分量；Step 3: Carry out empirical mode decomposition on the time series Δ[HbO₂ ](t) of oxyhemoglobin concentration changes obtained in step 2 and the time series Δ[HHb](t) of reduced hemoglobin concentration changes respectively, and obtain all The IMF component of步骤四、对步骤三获得的所有IMF分量进行希尔伯特变换，求取各个IMF分量的瞬时频率，将瞬时频率处于正常人呼吸频率和心脏跳动频率范围内的IMF分量剔除，以消除近红外脑功能检测时的生理扰动。Step 4. Perform Hilbert transform on all the IMF components obtained in step 3, obtain the instantaneous frequency of each IMF component, and remove the IMF components whose instantaneous frequency is within the range of normal human respiratory rate and heartbeat frequency to eliminate near-infrared Physiological perturbations during brain function testing.2.根据权利要求1所述的基于经验模态分解的近红外脑功能检测的扰动消除方法，其特征在于，步骤一所述的双波长光源发出的两种波长分别为λ₁＝750nm，λ₂＝830nm。2. the disturbance elimination method based on the near-infrared brain function detection of empirical mode decomposition according to claim 1, it is characterized in that, two kinds of wavelengths that the dual-wavelength light source described in step 1 sends are respectively λ₁ =750nm, λ₂ = 830nm.3.根据权利要求1所述的基于经验模态分解的近红外脑功能检测的扰动消除方法，其特征在于，步骤一中光密度变化量的时间序列

按如下公式获取：3. the disturbance elimination method based on the near-infrared brain function detection of empirical mode decomposition according to claim 1, is characterized in that, the time series of optical density variation in step 1

Obtain according to the following formula:$\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}{\mathrm{<span><span>OD</span>OD</span>}}_{{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\mathrm{<span><span>log</span>log</span>}{\mathrm{<span><span>I</span>I</span>}}_{\mathrm{<span><span>base</span>base</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span><span><span>/</span>/</span>{\mathrm{<span><span>I</span>I</span>}}_{\mathrm{<span><span>stim</span>stim</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>1</span>1</span>}}<span><span>)</span>)</span><span><span>,</span>,</span>$其中：I_base(λ₁)为探头光源的波长为λ₁时，大脑处于安静状态下时的出射光强，I_stim(λ₁)为探头光源的波长为λ₁时，大脑处于诱发激励时的出射光强，光密度变化量的时间序列

按如下公式获取：Among them: I_base (λ₁ ) is the outgoing light intensity when the brain is in a quiet state when the wavelength of the probe light source is λ₁ , and I_stim (λ₁ ) is when the wavelength of the probe light source is λ₁ and the brain is in evoked excitation The time series of the outgoing light intensity and optical density variation

Obtain according to the following formula:$\mathrm{<span><span>\&Delta;</span>\&Delta;</span>}{\mathrm{<span><span>OD</span>OD</span>}}_{{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\mathrm{<span><span>log</span>log</span>}{\mathrm{<span><span>I</span>I</span>}}_{\mathrm{<span><span>base</span>base</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span><span><span>/</span>/</span>{\mathrm{<span><span>I</span>I</span>}}_{\mathrm{<span><span>stim</span>stim</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>\&lambda;</span>\&lambda;</span>}}_{\mathrm{<span><span>2</span>2</span>}}<span><span>)</span>)</span><span><span>,</span>,</span>$其中：I_base(λ₂)为探头光源的波长为λ₂时，大脑处于安静状态下时的出射光强，I_stim(λ₂)为探头光源的波长为λ₂时，大脑处于诱发激励时的出射光强。Among them: I_base (λ₂ ) is the outgoing light intensity when the brain is in a quiet state when the wavelength of the probe light source is λ₂ , and I_stim (λ₂ ) is when the wavelength of the probe light source is λ₂ and the brain is in evoked excitation of the emitted light intensity.4.根据权利要求1所述的基于经验模态分解的近红外脑功能检测的扰动消除方法，其特征在于，步骤三中对氧合血红蛋白浓度变化量的时间序列Δ[HbO₂](t)和还原血红蛋白浓度变化量的时间序列Δ[HHb](t)进行经验模态分解的过程相同，以下将Δ[HbO₂](t)和Δ[HHb](t)统称为C_ij(t)，i为第几个IMF分量，i＝1，2，…，n，j为估计次数，初始化i＝1，j＝1，对时间序列C_ij(t)进行经验模态分解获取所有的IMF分量的获取过程：4. The method for eliminating disturbance based on near-infrared brain function detection based on empirical mode decomposition according to claim 1, characterized in that, in step 3, the time series Δ[HbO₂ ](t) of oxyhemoglobin concentration variation It is the same as the process of empirical mode decomposition of the time series Δ[HHb](t) of the reduced hemoglobin concentration variation. Hereinafter, Δ[HbO₂ ](t) and Δ[HHb](t) are collectively referred to as C_ij (t) , i is the number of IMF components, i=1, 2,..., n, j is the number of estimates, initialize i=1, j=1, perform empirical mode decomposition on time series C_ij (t) to obtain all IMFs Component acquisition process:步骤1、采用区域极值法确定寻找时间序列C_ij(t)的所有极大值和极小值，对获得的所有极大值、极小值分别用三次样条插值构建时间序列C_ij(t)的上包络线e_max(t)和下包络线e_min(t)；Step 1. Use the regional extremum method to determine and find all the maximum and minimum values of the time series C_ij (t), and use cubic spline interpolation to construct the time series C_ij ( t) the upper envelope e_max (t) and the lower envelope e_min (t);步骤2、获取上、下包络线的均值

Step 2. Obtain the mean value of the upper and lower envelopes

$\stackrel{<span><span>\&OverBar;</span>\&OverBar;</span>}{\mathrm{<span><span>e</span>e</span>}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span>}<span><span>=</span>=</span>\frac{{\mathrm{<span><span>e</span>e</span>}}_{\mathrm{<span><span>max</span>max</span>}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>+</span>+</span>{\mathrm{<span><span>e</span>e</span>}}_{\mathrm{<span><span>min</span>min</span>}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span>}{\mathrm{<span><span>2</span>2</span>}}<span><span>,</span>,</span>$步骤3、获取时间序列第i个IMF分量的第j次估计h_ij(t)：Step 3. Obtain the j-th estimate h_ij (t) of the i-th IMF component of the time series:${\mathrm{<span><span>h</span>h</span>}}_{\mathrm{<span><span>ij</span>ij</span>}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>=</span>=</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>ij</span>ij</span>}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>-</span>-</span>\stackrel{<span><span>\&OverBar;</span>\&OverBar;</span>}{\mathrm{<span><span>e</span>e</span>}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span>}<span><span>,</span>,</span>$步骤4、判断下式是否成立：

其中ε＞0，且充分接近0，Step 4. Determine whether the following formula is established:

Where ε>0, and sufficiently close to 0,判断结果为是，执行步骤5，If the judgment result is yes, go to step 5,判断结果为否，令j＝j+1，C_i(j+1)(t)＝h_ij(t)，并返回执行步骤1，If the judgment result is no, set j=j+1, C_i(j+1) (t)=h_ij (t), and return to step 1,步骤5、获取第i个IMF分量：C_i(t)＝h_ij(t)，并获得第i个残差：Step 5. Obtain the i-th IMF component: C_i (t)=h_ij (t), and obtain the i-th residual:r_i(t)＝C_ij(t)-h_ij(t)，r_i (t)=C_ij (t)-h_ij (t),步骤6、判断第i个残差r_i(t)是否为单调函数，Step 6. Determine whether the i-th residual r_i (t) is a monotone function,判断结果为否，令i＝i+1，j＝1，并返回执行步骤1，If the judgment result is no, let i=i+1, j=1, and return to execute step 1,判断结果为是，完成对时间序列C_ij(t)进行经验模态分解获取所有的IMF分量：C_i(t)。If the judgment result is yes, the empirical mode decomposition of the time series C_ij (t) is completed to obtain all IMF components: C_i (t).5.根据权利要求1所述的基于经验模态分解的近红外脑功能检测的扰动消除方法，其特征在于，步骤四中IMF分量的瞬时频率的获取方法为：5. the disturbance elimination method based on the near-infrared brain function detection of empirical mode decomposition according to claim 1, is characterized in that, the acquisition method of the instantaneous frequency of IMF component in the step 4 is:步骤41、获得IMF分量C_i(t)的希尔伯特变换y(t)：Step 41. Obtain the Hilbert transform y(t) of the IMF component C_i (t):$\mathrm{<span><span>y</span>the\; y</span>}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>=</span>=</span>\mathrm{<span><span>H</span>h</span>}<span><span>[</span>[</span>{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>i</span>i</span>}}<span><span>(</span>(</span>\mathrm{<span><span>t</span>t</span>}<span><span>)</span>)</span><span><span>]</span>]</span><span><span>=</span>=</span>\frac{\mathrm{<span><span>1</span>1</span>}}{\mathrm{<span><span>\&pi;</span>\&pi;</span>}}\mathrm{<span><span>P</span>P</span>}{<span><span>\&Integral;</span>\&Integral;</span>}_{<span><span>-</span>-</span><span><span>\&infin;</span>\&infin;</span>}^{<span><span>\&infin;</span>\&infin;</span>}\frac{{\mathrm{<span><span>C</span>C</span>}}_{\mathrm{<span><span>i</span>i</span>}}<span><span>(</span>(</span>{\mathrm{<span><span>t</span>t</span>}}^{<span><span>\&prime;</span>\&prime;</span>}<span><span>)</span>)</span>}{\mathrm{<span><span>t</span>t</span>}<span><span>-</span>-</span>{\mathrm{<span><span>t</span>t</span>}}^{<span><span>\&prime;</span>\&prime;</span>}}\mathrm{<span><span>d</span>d</span>}{\mathrm{<span><span>t</span>t</span>}}^{<span><span>\&prime;</span>\&prime;</span>}<span><span>,</span>,</span>$其中，P表示柯西主值，Among them, P represents the Cauchy principal value,步骤42、IMF分量C_i(t)的解析信号为z(t)＝C(t)+iy(t)＝a(t)exp[iθ(t)]，其中，a(t)为瞬时幅度，θ(t)为相位函数，Step 42, the analytical signal of the IMF component C_i (t) is z(t)=C(t)+iy(t)=a(t)exp[iθ(t)], where a(t) is the instantaneous amplitude , θ(t) is the phase function,步骤43、获取IMF分量C_i(t)的瞬时频率f(t)为：

Step 43, obtaining the instantaneous frequency f(t) of the IMF component C_i (t) is:

6.根据权利要求1所述的基于经验模态分解的近红外脑功能检测的扰动消除方法，其特征在于，正常人呼吸频率范围为0.15Hz～0.4Hz，心脏跳动频率范围为1.0Hz～1.7Hz。6. The disturbance elimination method of near-infrared brain function detection based on empirical mode decomposition according to claim 1, characterized in that the breathing frequency range of a normal person is 0.15 Hz to 0.4 Hz, and the heart beating frequency range is 1.0 Hz to 1.7 Hz. Hz.

Images