CN102288548A - Measuring method for improving ingredient detection accuracy of turbid medium - Google Patents

Abstract

The present invention relates to optical parameter measurements. In order to provide a measurement method with high detection accuracy and model applicability for the chemical components in the turbid medium, the technical solution adopted in the present invention is: firstly, the calibration set samples of the components to be measured should be designed so that the concentration changes in an equal gradient, and then through The double integrating sphere measurement system combines the inverse doubling method IAD to measure the absorption coefficient and scattering coefficient of the substance to be measured. According to a certain evaluation function, the substance to be measured is divided into absorption-based or scattering-based media. The absorption-dominated medium takes the absorption coefficient as a new substitute variable, and the scattering-dominated medium takes the scattering coefficient as a new substitute variable. The concentration matrix builds a linear calibration model, and finally uses this model to predict the component concentration of the unknown sample to be tested. The invention is mainly applied to the detection of chemical components in turbid media.

Description

A measurement method for improving the detection accuracy of components in turbid media

technical field

The invention relates to the field of optical parameter measurement in tissue optical research, in particular to a measurement method for improving the detection accuracy of components in turbid media.

Background technique

The research of the present invention belongs to the part of tissue optics, which mainly studies the law of light propagation in turbid medium and determines the distribution of light radiation energy in turbid medium under certain conditions. Therefore, we can apply the relevant theories of tissue optics to the detection of components in turbid media, study the interaction between light and media, and lay the foundation for further detection of the concentration of components in turbid media.

Tissue optics is a new discipline derived from the discipline of optics. It is a marginal discipline where optics and life science intersect and permeate each other. dissemination and distribution. The basic method is to establish a certain physical-mathematical model based on the optical characteristic parameters of the tissue (as shown in Table 1), and use analytical theories such as light transmission theory and numerical theories such as Monte Carlo method and adding doubling method, The distribution of light in the tissue is obtained by calculation or simulation, which provides a basis for the optical diagnosis and treatment of the tissue.

Table 1 Optical characteristic parameters of tissue

Optical measurement of biological tissue can be divided into two methods: in vitro measurement and in vivo measurement. The detection of optical parameters measured in vivo is more complex, and is more affected by the internal organization, and the accuracy of the measurement results is low. Therefore, in the current research, we often use in vitro measurement. Although in vitro measurement will bring some pain to patients in medical diagnosis, due to its high detection accuracy and mature and convenient measurement method, it is suitable for tissue optical measurement. has been widely applied.

Among the in vitro detection methods, integrating sphere technology is a widely used, relatively mature technology and relatively accurate measurement method. Integrating sphere technology is divided into single integrating sphere technology and double integrating sphere technology. Compared with single integrating sphere technology, double integrating sphere technology can effectively eliminate the background noise caused by the change of the position of the sample, and can be used without changing the position of the sample. In this case, the reflection and projection information of the sample can be obtained.

The double integrating sphere technique was first used to measure the optical parameters of myocardium and scleral tissue in 1990, and has now become an important tool for measuring the diffuse reflectance, diffuse transmittance and collimated transmittance of biological tissues. When measuring, the sample to be measured is placed between two integrating spheres, that is, when the light passes through the sample, reflection and transmission will occur, and the reflectance, diffuse transmittance and collimated transmittance of the sample to light can be obtained by the detector. And according to the theory of light transmission in tissue, the optical characteristic parameters of the sample are deduced: absorption coefficient μ _a , scattering coefficient μ _s and anisotropy factor g. The specific measurement principle is as follows:

The light intensity detected by the detector is proportional to the power P _d incident on the photodetector. In the linear working range of the photodetector, the output voltage V of the detector is proportional to the light intensity, that is, V=kP _d , k is a constant of proportionality, which depends on the characteristics of the detector. However, in actual measurement, due to the noise of the detector and the influence of external stray light, the background light V _n should also be considered, so the above formula can be rewritten as VV _n =kP _d .

Neglecting the absorption of the sphere and the loss of each port, the diffuse reflectance and diffuse transmittance can be defined as:

V _x = (V _{measure, x} −V _n )/(V ₀ /R _ref −V _pn )

In the formula: V _x represents the diffuse reflectance R _d and diffuse transmittance T _d , V _{measure, x} represents the voltage detected by the diffuse reflection and diffuse transmission detectors when there is a sample, V _n is the background noise, V ₀ is no sample Put the standard reflector on the light outlet, record the voltage detected by the diffuse reflection or diffuse transmission detector when the light is input into the integrating sphere, R _ref is the reflection coefficient of the standard reflector, V _pn is the case without sample Let the light pass through the integrating sphere to measure the voltage obtained by the diffuse reflection or diffuse transmission detector. The schematic diagram of the measurement system is shown in Figure 1.

Here, it is worth noting that the integrating sphere uses a relative measurement technique, that is, the diffuse reflectance, diffuse transmittance and collimated transmittance are all measured relative to a certain baseline, and the measured value is between 0-1. and cannot be greater than 1.

If the collimation performance of the laser is good, there will be V _pn ≈ V _n after the optical path is adjusted. In addition, in the case of eliminating the interference of background noise and adjusting the optical path, it can be considered that V _pn ≈V _n ≈0. Therefore, the diffuse reflectance, diffuse transmittance and collimated transmittance can be simplified as:

R _d = R _{measure, d} R _ref / R _{d_0} , T _d = T _{measure, d} R _ref / T _{d_0} , T _c = T _{measure, c} / T _{c_0}

When the light distribution in the turbid medium is known, the optical characteristic parameters of the turbid medium can be obtained by simulation, which is called the inverse problem of light propagation. Inverse Monte Carlo (Inverse Monte Carlo, IMC) and Inverse Doubling (Inverse Adding Doubling, IAD) are the two most commonly used methods for solving optical parameters. In the analysis and study of turbid media, it is a very important task to solve the optical parameters of the medium, which is the basis of the whole research process.

The IMC method is more accurate in calculating optical parameters than the IAD method, but its calculation time is longer. The IAD method is usually used to reverse the optical parameters of biological tissues. The IAD method has a fast calculation speed, and there is no limit to the albedo of the tissue and it can be flexible. It deals with the boundary conditions of biological tissues, thus greatly expanding its application range.

The IAD method has the following steps:

(1) Randomly generate the reflectance and transmittance values of light to biological tissue, and calculate the initial optical parameters

Firstly, the reduced albedo α′, the reduced optical depth τ′ and the reduced anisotropy factor g′ are respectively defined by the albedo α optical depth τ and the anisotropy factor g, namely

α′=α(1-g)/(1-αg), τ′=(1-αg)τ, g′=0 (1)

It can be obtained from the above formula: α=α'/(1-g+α'g), τ=τ'+α'τ'g/(1-g) (2)

In addition, the relationship between the reduced albedo α′ and the diffuse reflectance R _d , diffuse transmittance T _d and collimated transmittance T _c is:

α'=1-[(1-4R _d -T _d -T _c )/(1-T _d -T _c )] ² , (R _d /(1-T _d -T _c )<0.1) (3)

α'=1-(4/9)[(1-4R _d -T _d -T)/(1-T _d -T _c )] ² , (R _d /(1-T _d -T _c )≥0.1 ) (4)

The relationship between the reduced optical depth τ′ and the diffuse reflectance R _d , diffuse transmittance T _d and collimated transmittance T _c is:

${\mathrm{\&tau;}}^{\&prime;}=\frac{-\ln \left(0.05\right)\&CenterDot;\ln (T_d+T_c)}{\ln R_c},$ When R _d ≤ 0.1 (5)

${\mathrm{\&tau;}}^{\&prime;}=2^{1+5(R_d+T_d+T_c)},$ When _Rd > 0.1 (6)

The following relationship exists between the optical depth τ and the collimated transmittance _Tc :

${\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; =</span>\frac{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}$

In the formula, r ₁ and r ₂ represent the reflectivity of the front and rear surfaces of the measured sample, and the optical depth τ can be easily calculated from the above formula.

According to formulas (3)-(7), α', τ' and τ are substituted into formula (2), and α and g are obtained, and then calculated from (1) and the definition of albedo α and optical depth τ in Table 1 The optical parameters μ _a , μ _s and μ _s ′ are obtained, namely:

μ _a =(1-α′)τ′/d, μ _s =α′τ′/[d(1-g)], μ _s ′=α′τ′/d (8)

At this time, the preliminary absorption coefficient μ _a , scattering coefficient μ s , reduced scattering coefficient μ _{s ′} and anisotropy factor are obtained by _arbitrary diffuse reflectance R _d , diffuse transmittance T _d and collimated transmittance T _c Calculated value of g.

(2) From the estimated optical parameters obtained in the first step, the reflection value and transmission value are calculated by the multiplication method

The basic idea of the multiplication method is assuming that the reflectance and transmission values of incident light at a certain angle of a thin layer of sheet-like biological tissue are known, then for a sheet-like biological tissue twice the thickness and with the same characteristics, it is considered It can be divided into two identical slices, and then add the reflectance and transmission values of each slice of tissue. Therefore, for the calculation of the reflection and transmission values of any sheet-shaped biological tissue, the reflection and transmission values of a thin layer per unit thickness can be calculated by the known characteristics, and then the reflection and transmission values of tissues of any thickness can be obtained by the multiplication method.

(3) Comparing the calculated reflectance value and transmittance value with the actual measured reflectance value and transmittance value, correcting the optical parameters of the measured sample.

(4) Check whether the calculated value matches the measured value, if not, repeat the above steps until the two sets of values match the position, and the optical parameters at this time are the optical parameters of the measured sample.

Through these years of research, near-infrared spectroscopy has made great progress in the detection of chemical components, but has not been able to achieve effective breakthroughs in the detection of chemical components in turbid media.

Contents of the invention

In order to overcome the deficiencies in the prior art and provide a measurement method capable of accurately detecting the concentration of components in a turbid medium, in order to achieve the above object, the technical solution adopted in the present invention is: a measurement method for improving the detection accuracy of components in a turbid medium, comprising the following steps:

(1) A series of calibration sample sets C _i (i = 1, 2, ..., n) are measured by existing chemical measurement methods to establish concentration gradients for the components to be measured, i represents the sample number, and there are n parts in total;

(2) The diffuse reflectance R _d , diffuse transmittance T d , and collimated transmittance _{T c are} measured by a double integrating sphere measurement system;

(3) The absorption coefficient μ _a and the scattering coefficient μ _s are inversely calculated by the inverse multiplication method IAD;

对步骤(1)中的校正集样品进行评价；将被测成分分为以吸收为主的介质或以散射为主的介质；(4) Select evaluation function

Evaluate the calibration set samples in step (1); divide the measured components into absorption-based media or scattering-based media;

(5) According to the evaluation of the calibration sample set in step (4), the absorption coefficient is selected as the substitution variable for the medium mainly based on absorption, and the scattering coefficient is selected as the substitution variable for the medium mainly based on scattering;

(6) Establish a linear calibration model for the substitution variable selected in step (5), namely the absorption coefficient or scattering coefficient, and the calibration sample set C _i (i=1, 2, ..., n) in step (1);

(7) Using the model established in step (6) to measure the component concentration of the unknown sample to be tested.

来确定替代变量的选取，若k＞＞1，则选取散射系数作为替代变量，若k＜＜1，则选取吸收系数作为替代变量。The described response quantity transformation method passes the evaluation function

To determine the selection of surrogate variables, if k>>1, the scattering coefficient is selected as the surrogate variable, and if k<<1, the absorption coefficient is selected as the surrogate variable.

The linear relationship model C _i =a· _μi +b (i is the sample number) between the selected substitution variable and the concentration, according to C _i =a· _μi +b, calculate the linear coefficient a, b, Thus, the calculation formula C _x =a·μ _x +b for any unknown concentration is obtained, where x represents an unknown sample.

Compared with other existing technologies, the present invention has the following advantages:

First, through effective variable conversion, extract variables that are linearly related to concentration changes to replace absorbance and other variables that carry both absorption and scattering information, separate the effects of scattering and absorption, and establish new variables and the relationship between the concentration of the component The linear relationship model improves the accuracy of the measured component information extraction and the practicability of the model.

Second, the method for measuring the component concentration in the turbid medium based on the new variable transformation of the present invention provides a theoretical and technical basis for component detection in biological tissues, and can become a research method with application value.

Third, the double integrating sphere measuring system and the IAD algorithm used in the present invention to measure the optical parameters of the medium are all well-established technologies. After a large number of experiments and verifications, the accuracy is reliable and the measurement is convenient.

Fourth, the present invention has a wide range of applications, is applicable to various complicated turbid media, and has reliable precision.

Description of drawings

Fig. 1 is a schematic diagram of a measurement system in the prior art.

Fig. 2 is a schematic diagram of the measurement process of the present invention.

Figure 3 is the double integrating sphere system used in the present invention.

Detailed ways

In the detection and research of chemical components in turbid media, due to the existence of a large number of scattering particles in turbid media, they have strong scattering properties in addition to absorption properties, and the attenuation information when light passes through a sample not only includes The attenuation caused by the absorption of light by various chemical components also includes the attenuation caused by the scattering of light by scattering particles, and, in most cases, it is impossible to distinguish the contribution of absorption and scattering to the light attenuation information, which is the failure of the traditional scattering correction method the real reason.

In order to solve this problem, the present invention adopts the idea of transforming variables, that is, according to the characteristics of the components to be measured, select a suitable new substitution variable, such as absorption coefficient or scattering coefficient as a new variable, with the help of corresponding signal extraction methods and mathematical construction The model method is used to establish the relationship model between the substitution variable and the measured component concentration, and then quantify the component concentration in the turbid medium. In this method, the selection of surrogate variables is a key point. If the surrogate variables are selected incorrectly, more nonlinearities will be brought about, and vice versa, higher sensitivity and precision will be brought about. Therefore, we need to select an evaluation standard and establish an evaluation function to determine whether the measured substance is mainly absorption or scattering. For absorption-dominated components, the absorption coefficient is chosen as the surrogate variable, while for scattering-dominated components, the scattering coefficient is chosen as the surrogate variable.

For the establishment of the linear relationship model between the substitution variable and the concentration, a large number of experimental verifications have been carried out. The experimental results show that the linear relationship model has high prediction accuracy and good stability.

The present invention establishes a new method for measuring component concentrations in turbid media based on variable conversion, and the specific implementation steps are as follows:

1. A series of calibration sample sets C _i (i=1, 2, ..., n) are used to measure the components to be measured with existing chemical measurement methods and establish concentration gradients, where i represents the sample number, and there are n parts in total;

2. Measure the diffuse reflectance R _d , diffuse transmittance T _d and collimated transmittance T _c of the n samples by using a double integrating sphere measurement system;

3. Calculate the absorption coefficient μ _ai /scattering coefficient μ _si (i=1, 2, .

对该校正集样品进行评价，将被测成分分为以吸收为主或以散射为主的介质，若k＜＜1，则说明散射系数变化引起的浓度变化远小于吸收系数变化引起的浓度变化，也就是说该介质是以吸收为主的介质；若k＞＞1时，则说明散射系数变化引起的浓度变化远大于吸收系数变化引起的浓度变化，即该介质是以散射为主的介质；4, select certain evaluation function, select evaluation function among the present invention

Evaluate the calibration set sample, and divide the measured components into absorption-based or scattering-based media. If k<<1, it means that the concentration change caused by the change of the scattering coefficient is much smaller than the concentration change caused by the change of the absorption coefficient. , that is to say, the medium is an absorption-dominated medium; if k>>1, it means that the concentration change caused by the change of the scattering coefficient is much larger than the concentration change caused by the change of the absorption coefficient, that is, the medium is mainly a scattering medium ;

5. Here, we will use the method of variable conversion, that is, for the absorption-dominated medium, the absorption coefficient is selected as the new substitution variable, and for the scattering-dominated medium, the scattering coefficient is selected as the substitution variable;

6. Establish a linear calibration model C _i =a·μ _i +b for the selected new substitution variable, that is, the absorption coefficient μ _ai /scattering coefficient μ _si and the concentration C _i of the measured component in the calibration sample set measured by the standard method , (i=1, 2,..., n); find the value of a and b therefrom, thus obtain the relational expression of arbitrary unknown concentration C _x and substitution variable μ _x : C _x =a·μ _x +b, measurement process The schematic diagram is shown in Figure 2.

The double integrating sphere system adopted in the present invention is as accompanying drawing 3, and laser light source adopts DH-HN250 helium-neon laser of Daheng Science and Technology Company, output wavelength 623.8nm, and integrating sphere is the model that Sphere-Optics Company produces is SPH-12 integrating sphere , The reflectivity of the reflective coating on the inner wall is >95%, the diameter of the integrating sphere is: 304.8±0.1mm, the diameter of the light entrance of the reflection sphere is 6.0±0.1mm, and the diameter of the light exit of the transmission sphere is 12.7±0.1mm. The detector used was a Si detector from Hamamatsu, Japan, model G2592. The entire system is controlled based on the LABVIEW program, and can directly operate on the collected data, which improves the controllability of the system. The stability analysis results of the system are shown in Table 2.

Table 2 Stability analysis of double integrating sphere system

During the research process of this patent, typical turbid media such as milk and intralipid were mainly studied. By adopting the method of variable conversion, a linear relationship model between its substitution variable and concentration was finally established, thereby achieving the accuracy of component concentration detection in turbid media. Purpose.

This method can be used not only for typical turbid media such as traditional milk, but also for other complex turbid media, and further lays the foundation for improving the detection accuracy of components in biological tissues.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the method of the present invention, some improvements can also be made, and these improvements should also be regarded as protection scope of the present invention.

Claims (3)

1. A measuring method for improving component detection accuracy in a turbid medium, characterized in that it comprises the following steps: (1) A series of calibration sample sets C _i (i = 1, 2, ..., n) are measured by existing chemical measurement methods to establish concentration gradients for the components to be measured, i represents the sample number, and there are n parts in total; (2) The diffuse reflectance R _d , diffuse transmittance T d , and collimated transmittance _{T c are} measured by a double integrating sphere measurement system; (3) The absorption coefficient μ _a and the scattering coefficient μ _s are inversely calculated by the inverse multiplication method IAD; (4) Select evaluation function Evaluate the calibration set samples in step (1); divide the measured components into absorption-based media or scattering-based media; (5) According to the evaluation of the calibration sample set in step (4), the absorption coefficient is selected as the substitution variable for the medium mainly based on absorption, and the scattering coefficient is selected as the substitution variable for the medium mainly based on scattering; (6) Establish a linear calibration model for the substitution variable selected in step (5), namely the absorption coefficient or scattering coefficient, and the calibration sample set C _i (i=1, 2, ..., n) in step (1); (7) Using the model established in step (6) to measure the component concentration of the unknown sample to be tested. 2. The method according to claim 1, characterized in that, said response conversion method is passed through an evaluation function

To determine the selection of surrogate variables, if k>>1, the scattering coefficient is selected as the surrogate variable, and if k<<1, the absorption coefficient is selected as the surrogate variable. 3. The method according to claim 1, characterized in that, the linear relationship model C _i =a·μ _i +b between the selected substitution variable and concentration, i is the sample number, according to C _i =a · μ _i + b, calculate the linear coefficients a, b, and thus obtain the calculation formula of any unknown concentration C _x = a · μ _x + b, x represents the unknown sample.

Images