Inversion method of skin tissue optical property parameters based on frequency-optimized diffusion equation

Abstract

The invention discloses a method for inverting optical characteristic parameters of skin tissue based on frequency-optimized diffusion equation, which belongs to the technical field of spatial frequency domain imaging of biological tissue, and specifically relates to an optical characteristic parameter inversion method; the method comprises the following steps: setting skin tissue simulation sample parameters; setting values ​​and combinations of optical characteristic parameters; obtaining diffuse reflectance through Monte Carlo simulation; constructing a skin tissue diffusion equation group and inverting through a least squares algorithm; obtaining a variation law of inversion error through calculation and analysis; setting an initial value, measuring and obtaining the skin diffuse reflectance and inverting the optical characteristic parameters; obtaining an optimal spatial frequency sampling interval and sampling interval and an optimal optical characteristic parameter through iterative optimization of the total attenuation coefficient; compared with the method for inverting optical characteristic parameters of skin tissue based on the diffusion equation, the method of the invention has the beneficial effect of being able to obtain an optimal spatial frequency sampling interval and sampling interval, and effectively reducing the inversion error of the optical characteristic parameters.

Description

Technical Field

[0001] The invention relates to a method for inverting optical characteristic parameters of skin tissue based on frequency-optimized diffusion equation, which belongs to the technical field of spatial frequency domain imaging of biological tissue, and specifically relates to a method for inverting optical characteristic parameters. Background Art

[0002] Theory and practice show that light absorption and light scattering are common phenomena in the transmission of photons inside a medium. The absorption and scattering characteristics of a medium are mainly described by two optical characteristic parameters, namely the absorption coefficient and the reduced scattering coefficient, and the optical characteristic parameters are closely related to the geometric and physical properties of the medium. Therefore, it is an effective way to discover the physical and biological characteristics of the medium itself by detecting the optical characteristic parameters of the medium. At present, many detection technologies for optical characteristic parameters, especially optical detection technologies, have developed rapidly. Among them, spatial frequency domain imaging technology (SFDI) has become a hot spot and research frontier due to its many advantages such as non-contact and wide-field imaging, and is being tried in burn assessment, skin tissue assessment, tumor tissue detection, brain tissue monitoring, and agricultural product quality assessment.

[0003] The spatial frequency domain imaging technology first uses structured illumination with multiple spatial frequencies and wavelengths to measure the diffuse reflectance of the medium, and then reversely deduce the absorption coefficient and reduced scattering coefficient of the corresponding wavelength of the medium through the light propagation model and other means based on the diffuse reflectance, and then characterizes and interprets the characteristics in combination with the measured medium. Among them, the inversion of optical characteristic parameters is a key link that connects the previous and the next. It gives the optical characteristic parameters of the medium based on the diffuse reflectance, and then gives its physical, chemical or biomedical characterization and interpretation based on the medium. Not only that, the inversion process belongs to an inverse problem, which is essentially equivalent to an ill-posed problem that does not have a unique solution. Therefore, the inversion of optical characteristic parameters has become the key and difficulty of spatial frequency domain imaging technology.

[0004] At present, two approaches are commonly used for inversion: solving the diffusion equation by least squares and establishing a lookup table. The former is more mature and more universal, but is limited by model errors and model establishment conditions; the latter is simple and practical, but is limited by data completeness, frequency selectivity and resolution. For skin tissue evaluation, the optical property parameter inversion method based on the diffusion equation shows good usability. However, there is a problem of large inversion error due to model errors, and the size of the inversion error is directly affected by the spatial frequency sampling interval and sampling interval. Therefore, optimizing the spatial frequency sampling interval and sampling interval is expected to be an effective way to reduce the inversion error of skin tissue optical property parameters based on the diffusion equation. Summary of the invention

[0005] In order to reduce the inversion error of the skin tissue optical characteristic parameter inversion method based on the diffusion equation, the present invention discloses a skin tissue optical characteristic parameter inversion method based on the frequency optimized diffusion equation. Compared with the skin tissue optical characteristic parameter inversion method based on the diffusion equation, the method of the present invention establishes the optimal spatial frequency adoption interval and sampling interval expressed by the total attenuation coefficient through Monte Carlo simulation, and obtains the optimal spatial frequency adoption interval and sampling interval through iterative optimization of the total attenuation coefficient. The purpose of the present invention is to reduce the inversion error of the skin tissue optical characteristic parameters.

[0006] The object of the present invention is achieved in that:

[0007] The method for inverting the optical characteristic parameters of skin tissue based on the frequency-optimized diffusion equation is characterized by comprising the following steps:

[0008] Step a: The skin tissue simulation sample is set as a flat plate with infinite area. The thickness of the simulation sample is set to d = 2 cm, and the total number of grid lines in the depth z direction is set to n. z = 200, the total number of grid lines in the radius r direction is set to n l =300, the grid line spacing, i.e., the resolution, is set to Δ = 0.01 cm; according to the optical properties of skin tissue, the anisotropy coefficient in the physical parameters of the simulated sample is set to g = 0.9, and the refractive index is set to n 2 =1.4, and the refractive index of the surrounding medium is set to n 1 =1, the number of incident photons is set to m=10 6 , the wavelength of light is set to λ=633nm; these parameter setting values ​​are placed into the Monte Carlo simulation model of biological tissue light transmission to form a Monte Carlo simulation model of skin tissue light transmission, and the input of the Monte Carlo simulation model of skin tissue light transmission is the optical characteristic parameter of skin tissue (μ a ,μ′ s ) and the spatial frequency f of the streak light on the skin surface j , the output is the simulated value of diffuse reflectance R(μ a ,μ′ s ,f j ), the skin tissue optical characteristic parameters include the absorption coefficient μ a and the reduced scattering coefficient μ′ s ;

[0009] Step b: according to the actual optical characteristic parameters of skin tissue, the input absorption coefficient μ of the Monte Carlo simulation model of skin tissue light transmission is set when the incident light wavelength is 633 nm. a Value and reduction of scattering coefficient μ′ s See Table 1 for the values ​​and their combinations:

[0010] Table 1 Input optical characteristic parameter setting values ​​and their 30 combinations (cm -1 )

[0011]

[0012] The total attenuation coefficient μ in Table 1 tr for

[0013] μ tr =μ a +μ′ s

[0014] make

[0015] f j =γ×μ tr

[0016] Among them, f j With the total attenuation coefficient μ tr The unit is normalized and expressed as γ; according to the set values ​​in Table 1, and considering the premise of the diffuse approximation equation f j <0.33μ tr , set the starting spatial frequency f min The value range is 0μ tr -0.08μ tr , the value interval is 0.01μ tr , a total of 9 values, set the end spatial frequency f max The value range is 0.09μ tr -0.33μ tr , the value interval is 0.03μ tr , a total of 9 values, for each group of optical characteristic parameters (μ a ,μ′ s ) that is, each total attenuation coefficient μ tr 81 spatial frequency sampling intervals are formed. The spatial frequency sampling interval Δf is set at μ′ from the perspective of the minimum relative error of the optical characteristic parameter inversion. s =10cm -1 0.3cm -1 , 0.4cm in other cases -1 ;

[0017] Step c: for each group of optical characteristic parameters (μ a ,μ′ s ), in each spatial frequency sampling interval [f min ,f max ], the spatial frequency sampling point f is obtained with the spatial frequency sampling interval Δf as the step size j =f min +j×Δf, where j=0,1,2,...,J, J=floor[(fmax -f min ) / Δf], floor() is the rounding function, and the optical characteristic parameter (μ a ,μ′ s ) and the spatial frequency sampling point f j Insert the Monte Carlo simulation model of light transmission in skin tissue, run the Monte Carlo simulation program of light transmission in skin tissue, and obtain each group of optical properties (μ a ,μ′ s ) under the skin tissue simulation sample in each spatial frequency sampling interval [f min ,f max ] spatial frequency sampling point f j Simulated diffuse reflectance value R(μ a ,μ′ s ,f j );

[0018] Step d: Change the refractive index of the skin tissue simulation sample to n 2 =1.4 is inserted into the biological tissue diffusion approximation equation to obtain

[0019]

[0020] Where j = 0, 1, 2, ..., J, then we get a system of equations containing J + 1 equations, and R (μ a ,μ′ s ,f j ) is the simulated value of diffuse reflectance of skin tissue obtained by Monte Carlo simulation; the least squares algorithm is used to solve the equations to obtain the inversion simulation value of the absorption coefficient Sum of reduced scattering coefficient inversion simulation values

[0021] Step e: Calculate the inversion simulation value according to the following formula: and The relative error and for

[0022]

[0023]

[0024] Then we get 30 groups of simulation samples in 81 spatial frequency sampling intervals [f min ,f max ] under the inversion simulation value and Finally, the relative errors of all inversion simulation values ​​are compared and analyzed, and the optimal starting frequency f is obtained based on the principle of minimizing the relative error of the inversion simulation value. min and the optimal termination frequency fmax and μ′ s See Table 2 for the relationship between:

[0025] Table 2 Optimal starting frequency f min and the optimal termination frequency f max and μ′ s The relationship between (cm -1 )

[0026]

[0027] Step f: using a diffuse reflectance measuring device to obtain an initial value R of diffuse reflectance of skin tissue m1 (μ a ,μ′ s ,f j ), the spatial frequency sampling interval selects 81 spatial frequency sampling intervals [f min ,f max ] in the maximum interval [0,14.52cm -1 ] as the initial value, the spatial frequency interval is selected as 0.4cm -1 As the initial value, the initial value of the diffuse reflectance of the skin tissue R m1 (μ a ,μ′ s ,f j ) is inserted into the biological tissue diffusion approximation equation, and we get

[0028]

[0029] Then we get a system of 37 equations, and use the least squares algorithm to solve the system of equations to get the initial value of the absorption coefficient inversion measurement. The initial value of the inversion measurement of the reduced scattering coefficient And according to the inversion measurement initial value and Calculate the initial value of the total attenuation coefficient inversion measurement

[0030]

[0031] Step g: Invert the initial value and As μ′ in Table 2 s and μ tr The value of , the optimized spatial frequency sampling interval is in

[0032]

[0033] The spatial frequency sampling interval is optimized to The spatial frequency interval is optimized when Take 0.3cm -1 ,when Take 0.4cm -1 , the diffuse reflectance measurement value R of the skin tissue is obtained by using the diffuse reflectance measurement device at this spatial frequency sampling interval and interval m-new (μ a ,μ′ s ,f j ), the diffuse reflectance measurement value of skin tissue R m-new (μ a ,μ′ s ,f j ) is inserted into the biological tissue diffusion approximation equation, and we get

[0034]

[0035] Then we get a system of equations containing J+1 equations, and use the least squares algorithm to solve the system of equations to get the optimized absorption coefficient inversion measurement value and optimized reduced scattering coefficient inversion measurements And based on the optimized inversion measurement value and Calculate the optimized total attenuation coefficient inversion measurement value

[0036] Step h: Determine the iteration termination condition Is it true, where δ = 0.04 cm -1 is the iteration termination threshold; if it holds, the optimal inversion measurement value obtained at the end of the iteration is and If not, use and Replace step g with and And return to step g.

[0037] Beneficial effects:

[0038] Compared with the skin tissue optical characteristic parameter inversion method based on the diffusion equation, the method of the present invention can obtain the optimal spatial frequency sampling interval and sampling interval, and effectively reduce the optical characteristic parameter inversion error. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flow chart of the method of the present invention.

[0040] Figure 2 It is the relative inversion error of the absorption coefficient under different sampling intervals of spatial frequency.

[0041] Figure 3 It is the relative inversion error of the reduced scattering coefficient under different sampling intervals of spatial frequency. DETAILED DESCRIPTION

[0042] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings.

[0043] The present invention is based on the frequency optimization diffusion equation skin tissue optical characteristic parameter inversion method, the specific process is as follows Figure 1 As shown, the following steps are included:

[0044] Step a: The skin tissue simulation sample is set as a flat plate with infinite area. The thickness of the simulation sample is set to d = 2 cm, and the total number of grid lines in the depth z direction is set to n. z = 200, the total number of grid lines in the radius r direction is set to n l =300, the grid line spacing, i.e., the resolution, is set to Δ = 0.01 cm; according to the optical properties of skin tissue, the anisotropy coefficient in the physical parameters of the simulated sample is set to g = 0.9, and the refractive index is set to n 2 =1.4, and the refractive index of the surrounding medium is set to n 1 =1, the number of incident photons is set to m=10 6 , the wavelength of light is set to λ=633nm; these parameter setting values ​​are placed into the Monte Carlo simulation model of biological tissue light transmission to form a Monte Carlo simulation model of skin tissue light transmission, and the input of the Monte Carlo simulation model of skin tissue light transmission is the optical characteristic parameter of skin tissue (μ a ,μ′ s ) and the spatial frequency f of the streak light on the skin surface j , the output is the simulated value of diffuse reflectance R(μ a ,μ′ s ,f j ), the skin tissue optical characteristic parameters include the absorption coefficient μ a and the reduced scattering coefficient μ′ s ;

[0045] Step b: according to the actual optical characteristic parameters of skin tissue, the input absorption coefficient μ of the Monte Carlo simulation model of skin tissue light transmission is set when the incident light wavelength is 633 nm. a Value and reduction of scattering coefficient μ′ s See Table 1 for the values ​​and their combinations:

[0046] Table 1 Input optical characteristic parameter setting values ​​and their 30 combinations (cm -1 )

[0047]

[0048] The total attenuation coefficient μ in Table 1 tr for

[0049] μ tr=μ a +μ′ s

[0050] make

[0051] f j =γ×μ tr

[0052] Among them, f j With the total attenuation coefficient μ tr The unit is normalized and expressed as γ; according to the set values ​​in Table 1, and considering the premise of the diffuse approximation equation f j <0.33μ tr , set the starting spatial frequency f min The value range is 0μ tr -0.08μ tr , the value interval is 0.01μ tr , a total of 9 values, set the end spatial frequency f max The value range is 0.09μ tr -0.33μ tr , the value interval is 0.03μ tr , a total of 9 values, for each group of optical characteristic parameters (μ a ,μ′ s ) that is, each total attenuation coefficient μ tr 81 spatial frequency sampling intervals are formed. The spatial frequency sampling interval Δf is set at μ′ from the perspective of the minimum relative error of the optical characteristic parameter inversion. s =10cm -1 0.3cm -1 , 0.4cm in other cases -1 ;

[0053] Step c: for each group of optical characteristic parameters (μ a ,μ′ s ), in each spatial frequency sampling interval [f min ,f max ], the spatial frequency sampling point f is obtained with the spatial frequency sampling interval Δf as the step size j =f min +j×Δf, where j=0,1,2,...,J, J=floor[(f max -f min ) / Δf], floor() is the rounding function, and the optical characteristic parameter (μ a ,μ′ s ) and the spatial frequency sampling point f j Insert the Monte Carlo simulation model of light transmission in skin tissue, run the Monte Carlo simulation program of light transmission in skin tissue, and obtain each group of optical properties (μ a,μ′ s ) under the skin tissue simulation sample in each spatial frequency sampling interval [f min ,f max ] spatial frequency sampling point f j Simulated diffuse reflectance value R(μ a ,μ′ s ,f j );

[0054] Step d: Change the refractive index of the skin tissue simulation sample to n 2 =1.4 is inserted into the biological tissue diffusion approximation equation to obtain

[0055]

[0056] Where j = 0, 1, 2, ..., J, then we get a system of equations containing J + 1 equations, and R (μ a ,μ′ s ,f j ) is the simulated value of diffuse reflectance of skin tissue obtained by Monte Carlo simulation; the least squares algorithm is used to solve the equations to obtain the inversion simulation value of the absorption coefficient Sum of reduced scattering coefficient inversion simulation values

[0057] Step e: Calculate the inversion simulation value according to the following formula: and The relative error and for

[0058]

[0059]

[0060] Then we get 30 groups of simulation samples in 81 spatial frequency sampling intervals [f min ,f max ] under the inversion simulation value and Finally, the relative errors of all inversion simulation values ​​are compared and analyzed, and the optimal starting frequency f is obtained based on the principle of minimizing the relative error of the inversion simulation value. min and the optimal termination frequency f max and μ′ s See Table 2 for the relationship between:

[0061] Table 2 Optimal starting frequency f min and the optimal termination frequency f max and μ′ s The relationship between (cm -1 )

[0062]

[0063] Step f: using a diffuse reflectance measuring device to obtain an initial value R of diffuse reflectance of skin tissue m1 (μ a ,μ′ s ,f j ), the spatial frequency sampling interval selects 81 spatial frequency sampling intervals [f min ,f max ] in the maximum interval [0,14.52cm -1 ] as the initial value, the spatial frequency interval is selected as 0.4cm -1 As the initial value, the initial value of the diffuse reflectance of the skin tissue R m1 (μ a ,μ′ s ,f j ) is inserted into the biological tissue diffusion approximation equation, and we get

[0064]

[0065] Then we get a system of 37 equations, and use the least squares algorithm to solve the system of equations to get the initial value of the absorption coefficient inversion measurement. The initial value of the inversion measurement of the reduced scattering coefficient And according to the inversion measurement initial value and Calculate the initial value of the total attenuation coefficient inversion measurement

[0066]

[0067] Step g: Invert the initial value and As μ′ in Table 2 s and μ tr The value of , the optimized spatial frequency sampling interval is in

[0068]

[0069] The spatial frequency sampling interval is optimized to The spatial frequency interval is optimized when Take 0.3cm -1 ,when Take 0.4cm -1 , the diffuse reflectance measurement value R of the skin tissue is obtained by using the diffuse reflectance measurement device at this spatial frequency sampling interval and interval m-new (μ a ,μ′ s ,f j ), the diffuse reflectance measurement value of skin tissue Rm-new (μ a ,μ′ s ′,f j ) is inserted into the biological tissue diffusion approximation equation, and we get

[0070]

[0071] Then we get a system of equations containing J+1 equations, and use the least squares algorithm to solve the system of equations to get the optimized absorption coefficient inversion measurement value and optimized reduced scattering coefficient inversion measurements And based on the optimized inversion measurement value and Calculate the optimized total attenuation coefficient inversion measurement value

[0072] Step h: Determine the iteration termination condition Is it true, where δ = 0.04 cm -1 is the iteration termination threshold; if it holds, the optimal inversion measurement value obtained at the end of the iteration is and If not, use and Replace step g with and And return to step g.

[0073] It should be noted that the technical field corresponding to the technical solution of the present invention is the technical field of spatial frequency domain imaging of biological tissues. As far as those skilled in the art are concerned, the diffuse reflectance measurement device in step f of the method of the present invention can be selected and applied by those skilled in the art based on their professional knowledge, and there is no need for specific principle explanation and instrument equipment description in the present invention. The Monte Carlo simulation model of biological tissue light transmission in step a, the biological tissue diffuse approximate equation in step d and the least squares solution algorithm for the equation group are well-known mature models and algorithms in the art, and those skilled in the art can fully implement them independently, and the present invention has been fully disclosed.

[0074] The following analysis illustrates that the method for inverting the optical characteristic parameters of skin tissue based on the frequency-optimized diffusion equation of the present invention has the technical advantage of smaller inversion error in principle compared with the method for inverting the optical characteristic parameters of skin tissue based on the diffusion equation.

[0075] The inversion method of skin tissue optical property parameters based on the diffusion equation has the problem of large inversion error. The use of limited spatial frequency sampling interval and sampling interval becomes a factor that causes the inversion error, and different frequency sampling intervals and sampling points lead to inversion errors of different sizes.

[0076] The method proposed in the present invention provides the optimal spatial frequency sampling interval and sampling interval under different conditions through Monte Carlo simulation of skin tissue before measurement, and obtains the optimal spatial frequency sampling interval and sampling interval through iterative optimization of the total attenuation coefficient during measurement, which can effectively reduce the inversion error of the optical characteristic parameters of skin tissue based on the diffusion equation.

[0077] The following simulation experiments are used to prove that the method of the present invention has a technical advantage of small inversion error compared with the skin tissue optical characteristic parameter inversion method based on the diffusion equation.

[0078] The experimental process and results are as follows:

[0079] For the 30 groups of skin tissue simulation samples in Table 1, the simulated values ​​of diffuse reflectance R (μ a ,μ′ s ,f j ) as the diffuse reflectance measurement value, first use the method of this invention to invert 30 groups of absorption coefficients and reduced scattering coefficients, then use the skin tissue optical characteristic parameter inversion method based on the diffusion equation to invert 30 groups of absorption coefficients and reduced scattering coefficients, and then use the optical characteristic parameters (μ a ,μ′ s ) as the true value, the relative error average MRE, root mean square error RMSE and absolute error maximum MAX of the optical characteristic parameter inversion in each inversion method are calculated. The experimental results are shown in Table 3:

[0080] Table 3 Comparison of inversion errors of optical characteristic parameters of 30 groups of skin tissue simulation samples

[0081]

[0082] It can be seen from Table 3 that compared with the skin tissue optical characteristic parameter inversion method based on the diffusion equation, the inversion μ a The MRE, RMSE and MAX of the inverted μ′ were reduced by 24.47%, 35.60% and 38.61% respectively. s The MRE, RMSE and MAX of the proposed method were reduced by 18.24%, 26.10% and 28.19% respectively, which verified that the method of the present invention can effectively reduce the inversion error of optical characteristic parameters.

Claims

1. Inversion method of skin tissue optical property parameters based on frequency optimized diffusion equation, It is characterized in that The following steps are involved: Step a: The skin tissue simulation sample is set as a flat plate with infinite area. The thickness of the simulation sample is set to d = 2 cm, and the total number of grid lines in the depth z direction is set to n. z = 200, the total number of grid lines in the radius r direction is set to n l =300, the grid line spacing, i.e., the resolution, is set to Δ = 0.01 cm; according to the optical properties of skin tissue, the anisotropy coefficient in the physical parameters of the simulated sample is set to g = 0.9, and the refractive index is set to n 2 =1.4, and the refractive index of the surrounding medium is set to n 1 =1, the number of incident photons is set to m=10 6 , the wavelength of light is set to λ=633nm; these parameter setting values ​​are placed into the Monte Carlo simulation model of biological tissue light transmission to form a Monte Carlo simulation model of skin tissue light transmission, and the input of the Monte Carlo simulation model of skin tissue light transmission is the optical characteristic parameter of skin tissue (μ a ,μ′ s ) and the spatial frequency f of the streak light on the skin surface j , the output is the simulated value of diffuse reflectance R(μ a ,μ′ s ,f j ), the skin tissue optical characteristic parameters include the absorption coefficient μ a and the reduced scattering coefficient μ′ s ; Step b: according to the actual optical characteristic parameters of skin tissue, the input absorption coefficient μ of the Monte Carlo simulation model of skin tissue light transmission is set when the incident light wavelength is 633 nm. a Value and reduction of scattering coefficient μ′ s See Table 1 for the values ​​and their combinations: Table 1 Input optical characteristic parameter setting values ​​and their 30 combinations (cm -1 ) The total attenuation coefficient μ in Table 1 tr for m tr =μ a +μ′ s make f j =γ×μ tr Among them, f j With the total attenuation coefficient μ tr The unit is normalized and expressed as γ; according to the set values ​​in Table 1, and considering the premise of the diffuse approximation equation f j <0.33μ tr , set the starting spatial frequency f min The value range is 0μ tr -0.08μ tr , the value interval is 0.01μ tr , a total of 9 values, set the end spatial frequency f max The value range is 0.09μ tr -0.33μ tr , the value interval is 0.03μ tr , a total of 9 values, for each group of optical characteristic parameters (μ a ,μ′ s ) that is, each total attenuation coefficient μ tr 81 spatial frequency sampling intervals are formed. The spatial frequency sampling interval Δf is set at μ′ from the perspective of the minimum relative error of the optical characteristic parameter inversion. s =10cm -1 0.3cm -1 , 0.4cm in other cases -1 ; Step c: for each group of optical characteristic parameters (μ a ,μ′ s ), in each spatial frequency sampling interval [f min ,f max ], the spatial frequency sampling point f is obtained with the spatial frequency sampling interval Δf as the step size j =f min +j×Δf, where j=0,1,2,…,J, J=floor[(f max -f min ) / Δf], floor() is the rounding function, and the optical characteristic parameter (μ a ,μ′ s ) and the spatial frequency sampling point f j Insert the Monte Carlo simulation model of light transmission in skin tissue, run the Monte Carlo simulation program of light transmission in skin tissue, and obtain each group of optical properties (μ a ,μ′ s ) under the skin tissue simulation sample in each spatial frequency sampling interval [f min ,f max ] spatial frequency sampling point f j Simulated diffuse reflectance value R(μ a ,μ′ s ,f j ); Step d: Change the refractive index of the skin tissue simulation sample to n 2 =1.4 is inserted into the biological tissue diffusion approximation equation to obtain Where j = 0, 1, 2, ..., J, then we get a system of equations containing J + 1 equations, and R (μ a ,μ′ s ,f j ) is the simulated value of diffuse reflectance of skin tissue obtained by Monte Carlo simulation; the least squares algorithm is used to solve the equations to obtain the inversion simulation value of the absorption coefficient Sum of reduced scattering coefficient inversion simulation values Step e: Calculate the inversion simulation value according to the following formula: and The relative error and for Then we get 30 groups of simulation samples in 81 spatial frequency sampling intervals [f min ,f max ] under the inversion simulation value and Finally, the relative errors of all inversion simulation values ​​are compared and analyzed, and the optimal starting frequency f is obtained based on the principle of minimizing the relative error of the inversion simulation value. min and the optimal termination frequency f max and μ′ s See Table 2 for the relationship between: Table 2 Optimal starting frequency f min and the optimal termination frequency f max and μ′ s The relationship between (cm -1 ) Step f: using a diffuse reflectance measuring device to obtain an initial value R of diffuse reflectance of skin tissue m1 (μ a ,μ′ s ,f j ), the spatial frequency sampling interval selects 81 spatial frequency sampling intervals [f min ,f max ] in the maximum interval [0,14.52cm -1 ] as the initial value, the spatial frequency interval is selected as 0.4cm -1 As the initial value, the initial value of the diffuse reflectance of the skin tissue R m1 (μ a ,μ′ s ,f j ) is inserted into the biological tissue diffusion approximation equation, and we get Then we get a system of 37 equations, and use the least squares algorithm to solve the system of equations to get the initial value of the absorption coefficient inversion measurement. The initial value of the inversion measurement of the reduced scattering coefficient And according to the inversion measurement initial value and Calculate the initial value of the total attenuation coefficient inversion measurement Step g: Invert the initial value and As μ′ in Table 2 s and μ tr The value of , the optimized spatial frequency sampling interval is in The spatial frequency sampling interval is optimized to The spatial frequency interval is optimized when Take 0.3cm -1 ,when Take 0.4cm -1 , the diffuse reflectance measurement value R of the skin tissue is obtained by using the diffuse reflectance measurement device at this spatial frequency sampling interval and interval m-new (μ a ,μ′ s ,f j ), the diffuse reflectance measurement value of skin tissue R m-new (μ a ,μ′ s ,f j ) is inserted into the biological tissue diffusion approximation equation, and we get Then we get a system of equations containing J+1 equations, and use the least squares algorithm to solve the system of equations to get the optimized absorption coefficient inversion measurement value and optimized reduced scattering coefficient inversion measurements And based on the optimized inversion measurement value and Calculate the optimized total attenuation coefficient inversion measurement value Step h: Determine the iteration termination condition Is it true, where δ = 0.04 cm -1 is the iteration termination threshold; If it holds true, the optimal inversion measurement value obtained at the end of the iteration is and If not, use and Replace step g with and And return to step g.

Images