CN103767686B - Method for positioning bioluminescence imaging light sources in small animal - Google Patents

Abstract

The invention discloses a light source positioning method for small animal bioluminescent imaging, which comprises the following steps: using a quantitative optical molecular tomography device and a finite element method to construct the relationship between a small animal body surface measurement data vector and the distribution of an unknown light source in the body; using algebraic iteration to reconstruct Methods The light source distribution in the small animal body was calculated; the threshold was determined according to the sparsity, and the light source distribution obtained by the algebraic iterative reconstruction method was used to correct the threshold; after multiple cycles, the light source distribution in the small animal body was finally obtained. The advantage of the present invention is that it is not necessary to add an _l0 regularization term in the mathematical model of the reconstruction problem, and it is not necessary to use the _l1 norm or the lp (0< _p <1) norm to carry out the _l0 norm Instead, the sparseness is directly used to correct the light source distribution obtained by the algebraic iterative reconstruction method. Since the norm approximation in the prior art is not used, the method of the present invention improves the light source positioning accuracy in small animals.

Description

A light source positioning method for small animal bioluminescence imaging

technical field

The invention relates to an imaging light source positioning method, in particular to a small animal bioluminescent imaging light source positioning method, belonging to the field of optical imaging.

Background technique

Bioluminescent imaging technology uses luciferase gene to mark cells or DNA, and uses semiconductor cooling CCD camera to collect optical signals, which can directly monitor cell activity and gene behavior in living organisms.

Bioluminescent imaging technology can also observe biological processes such as tumor growth and metastasis, infectious disease development, and specific gene expression in living animals.

Bioluminescent imaging technology has the characteristics of no ionizing radiation, high sensitivity, and low cost, and is widely used in biological research.

One of the core issues of bioluminescence imaging technology is the location of bioluminescent light sources in small animals, which can be obtained by reconstructing the distribution of light sources in vivo based on the measured surface fluorescence signals of small animals. Since the number of measured data is less than the number of unknowns, the solution to the bioluminescence imaging reconstruction problem is not unique. In order to obtain a solution close to the real distribution of the light source, a regularization term can be added to the objective function of the reconstruction problem. Considering the sparse distribution of light sources in small animals, the researchers proposed to use the l ₀ norm to constrain the added regularization term. Mathematically, the l ₀ norm regularization problem is difficult to solve. In practice, the l ₁ norm or the l _p (0<p<1) norm is usually used to approximate the l ₀ norm.

Chinese invention patent "A Bioluminescence Tomography Reconstruction Method", application number 201310259527.1, application date 20160626, publication date 20130904, discloses a bioluminescence tomography reconstruction method, adding l _0.5 regularization term to the objective function of the reconstruction problem, and The l _0.5 regularized objective function is transformed into a reweighted l ₁ regularized minimization problem by using the weighted interior point method, and then the minimization problem is solved by using the interior point method to obtain the three-dimensional positioning and quantitative information of the fluorescent light source in the organism. Due to the approximation of the l ₀ norm, the reconstruction error must be introduced, resulting in inaccurate positioning.

Contents of the invention

In order to solve the deficiencies of the prior art, the purpose of the present invention is to provide a method for locating the light source for bioluminescence imaging of small animals. The method first uses the algebraic iterative reconstruction (ART) method to calculate the light source distribution, and then uses the threshold value to correct the above light source distribution. Make the sparsity of the corrected light source distribution meet the given conditions, then use the corrected light source distribution as the initial value, continue to use the ART method to calculate the new light source distribution, and repeat it many times until the small animal object calculated according to the light source distribution When the error between the surface fluorescence signal and the fluorescence signal detected by the CCD is less than a given error value, the calculation ends, and finally the position of the light source is calculated according to the distribution of the light source, so as to realize the accurate positioning of the light source in the small animal body.

In order to achieve the above object, the present invention adopts the following technical solutions:

A method for positioning a light source for bioluminescence imaging of small animals, comprising the following steps:

(1) Acquisition of optical signals and internal structure information on the body surface of small animals

1.a Use quantitative optical molecular tomography to obtain two-dimensional bioluminescence images on the surface of small animals and three-dimensional computed tomography images of internal structures;

1.b Arrange the collected bioluminescent images into data vectors, and use the finite element method to construct the relationship between the data vectors and the distribution of unknown light sources in the body, as follows:

y=Ax+n (1)

In the formula, y is obtained from the bioluminescence image, and the size is M rows and 1 column,

A is the coefficient matrix obtained from computed tomography images, the size of which is M rows and N columns,

x is the unknown light source distribution in the small animal body, the size is N rows and 1 column,

n is noise, the size is M rows and 1 column;

(2) Set the initial value

Set the initial light source distribution x, the initial threshold β, and the initial sparsity Among them, x≥0, β≥1,

(3) Iteratively update the light source distribution

3.a Take the first row of the coefficient matrix, denoted as A ₁ , take the first element of the data vector Y, denoted as y ₁ , and calculate the increment:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\frac{{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}{{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

In the formula, is the square of the l ₂ norm of A ₁ ,

γ is the weight, γ<1;

3.b Utilize the above increment △ to update x, and the updated x is denoted by x', which has:

x'=x+△ (3)

3.c Substitute the x' obtained in step 3.b into formula (2), take the second row A ₂ of the coefficient matrix A and the second element y ₂ of the vector y, calculate the new increment △, and continue to use the formula ( 3) Update x; until the increment is calculated using the Mth row of the coefficient matrix A and the Mth element of the vector y and the update of x is completed, the light source distribution after the update is recorded as x _new ;

(4) Threshold correction light source distribution

4.a Take the largest element of x _new obtained in step 3.c, record it as x_max, and calculate $\mathrm{\&alpha;}=\frac{x\_\max }{\mathrm{\&beta;}};$

4.b Compare the first element to the Nth element of x _new with α in step 4.a in turn. If the aforementioned element is smaller than α, then set the aforementioned element to zero to obtain the corrected light source distribution, which is denoted as

4.c Use the following formula to calculate the above corrected light source distribution The sparsity of:

$\mathrm{<span\; class="notranslate">\; \&psi;</span>}<span\; class="notranslate">\; =</span>\frac{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; new</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; new</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

In the formula, respectively The l ₁ norm and l ₂ norm of ;

4.d Compare the sparsity ψ obtained in step 4.c with the initial sparsity To compare, if Then change the threshold β and return to steps 4.a, 4.b and 4.c for calculation and judgment; if Then the obtained threshold value and the corrected light source distribution are recorded as and Execute step (5); ε is the error;

(5) Calculation error and judgment stop condition

According to the corrected light source distribution in step 4.d calculate and will Compare with y in formula (1), if then get from step 4.d and As the initial threshold and the initial light source distribution, ie and Return to step (3) and step (4), iteratively update the light source distribution and threshold correction light source distribution again; if As the final light source distribution, denoted as x _opt , perform step (6);

(6) Light source positioning

According to the final light source distribution x _opt obtained in step (5), the maximum element position of the aforementioned final light source distribution is found to complete the light source positioning.

The aforementioned small animal bioluminescence imaging light source positioning method is characterized in that, in step (2), when setting the initial value, x=0, β=2,

The aforementioned light source positioning method for small animal bioluminescence imaging is characterized in that, in step (3), γ=0.25 when iteratively updating the light source distribution and calculating the increment.

The aforementioned light source positioning method for small animal bioluminescence imaging is characterized in that, in step (4) and step (5), the error ε=1e ^-6 .

The aforementioned light source positioning method for bioluminescent imaging of small animals is characterized in that when using a quantitative optical molecular tomography device to acquire bioluminescent images of the body surface of small animals and computed tomographic images of internal structures, the posture of the small animals remains unchanged.

The benefit of the present invention is that: the light source positioning method of the present invention does not need to add an _l0 regularization term in the mathematical model of the reconstruction problem, and does not need to use the _l1 norm or the lp (0< _p <1) norm The l ₀ norm is approximated, but the light source distribution obtained by the ART method is directly used to correct the sparsity, so as to realize the positioning of the bioluminescent light source in the small animal body. Since the norm approximation in the prior art is not used, The method of the invention improves the positioning accuracy of the light source in the small animal body, and can be used in research fields such as early tumor detection and treatment tracking.

Description of drawings

FIG. 1 is a flow chart of the method for positioning a light source for small animal bioluminescence imaging of the present invention.

detailed description

The present invention will be specifically introduced below in conjunction with the accompanying drawings and specific embodiments.

Referring to Fig. 1, the positioning method of the light source for small animal bioluminescence imaging of the present invention comprises the following steps:

1. Anesthetizing and immobilizing small animals

Anesthetize a small animal with a bioluminescent light source in its body, and fix the limbs, head and tail of the small animal on the sample holder.

2. Obtain the optical signal and internal structure information of the small animal body surface

2.a In the darkroom, using the quantitative optical molecular tomography device disclosed in Chinese invention patent ZL201010173473.3, first collect the fluorescent signals on the body surface of small animals, and collect them every 90 degrees. Four fluorescent images can be collected when the small animal rotates once. Image: Under the premise that the posture of the small animal remains unchanged, the projection data of the computerized tomography imaging is collected once every 1 degree, and the projection data is collected 360 times when the small animal rotates a circle, and the three-dimensional computed tomography of the small animal is obtained by using the back projection algorithm image.

2.b Arrange the four collected fluorescence images into data vectors, use the professional software Amira to segment the three-dimensional computed tomography images of small animals, and use the diffusion equation to model the light transmission process in small animals. The meta-method can construct the relationship between the data vector of the fluorescent signal on the surface of the small animal and the unknown light source in the small animal, as follows:

y=Ax+n (1)

In the formula, y is obtained from the bioluminescence image, and the size is M rows and 1 column,

A is the coefficient matrix obtained from computed tomography images, the size of which is M rows and N columns,

x is the unknown light source distribution in the small animal body, the size is N rows and 1 column,

n is noise, the size is M rows and 1 column.

3. Set the initial value

Set the initial light source distribution x, the initial threshold β, and the initial sparsity Among them, x≥0, β≥1, Preferably, x=0, β=2,

4. Iteratively update the light source distribution

4.a Take the first row of the coefficient matrix, denoted as A ₁ , take the first element of the data vector Y, denoted as y ₁ , and calculate the increment:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\frac{{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}{{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

In the formula, is the square of the l ₂ norm of A ₁ ,

γ is a weight, γ<1, preferably, γ=0.25.

4.b Utilize the above increment △ to update x, and the updated x is denoted by x', which has:

x'=x+△ (3)

4.c Substitute x' obtained in step 4.b into formula (2), take the second row A ₂ of the coefficient matrix A and the second element y ₂ of the vector y, calculate the new increment △, and continue to use the formula ( 3) Update x; until the increment is calculated using the Mth row of the coefficient matrix A and the Mth element of the vector y and the update of x is completed, the light source distribution after the update is recorded as x _new .

5. Threshold correction light source distribution

5.a Take the largest element of x _new obtained in step 4.c, record it as x_max, and calculate $\mathrm{\&alpha;}=\frac{x\_\max }{\mathrm{\&beta;}}.$

5.b Compare the first element of x _new with α in step 5.a, if the first element is less than α, set the first element to zero; if the first element greater than or equal to α, the first element remains unchanged.

Compare sequentially until the last element of x _new to obtain the corrected light source distribution, which is denoted as

5.c Use the following formula to calculate the above-mentioned corrected light source distribution The sparsity of:

$\mathrm{<span\; class="notranslate">\; \&psi;</span>}<span\; class="notranslate">\; =</span>\frac{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; new</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; new</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

In the formula, respectively The l ₁ norm and l ₂ norm of The less 1 in , the closer the sparsity ψ is to 1, When only one element is 1, ψ=1.

5.d Compare the sparsity ψ obtained in step 5.c with the initial sparsity To compare, if Then change the threshold β and return to steps 5.a, 5.b and 5.c for calculation and judgment; if Then the obtained threshold value and the corrected light source distribution are recorded as and Go to step 6.

ε is an error, which is a small amount, preferably, ε=1e ^-6 .

6. Calculate the error and judge the stop condition

According to the corrected light source distribution in step 5.d calculate and will Compared with the small animal body surface fluorescence signal data vector y in formula (1), if then get from step 5.d and As the initial threshold and initial light source distribution, ie and Return to step 4 and step 5, iteratively update the light source distribution and threshold value correction light source distribution again; if As the final light source distribution, denoted as x _opt , go to step 7.

7. Light source positioning

According to the final light source distribution x _opt obtained in step 6, the maximum element position of the final light source distribution is found to complete the light source positioning.

The light source positioning method of the present invention does not need to add an _l0 regularization term in the mathematical model of the reconstruction problem, and does not need to use the _l1 norm or the lp (0< _p <1) norm to approximate the _l0 norm , but directly uses the sparsity to correct the light source distribution obtained by the ART method. Since the norm approximation in the prior art is not used, the method of the present invention improves the positioning accuracy of the light source in the small animal body.

The light source positioning method of the present invention can be used in research fields such as early tumor detection and treatment tracking.

It should be noted that the above embodiments do not limit the present invention in any form, and all technical solutions obtained by means of equivalent replacement or equivalent transformation fall within the protection scope of the present invention.

Claims (5)

1. A method for positioning a light source for small animal bioluminescence imaging, comprising the following steps: 1. Acquisition of small animal surface optical signals and internal structure information: 1.a Use quantitative optical molecular tomography to obtain two-dimensional bioluminescence images on the surface of small animals and three-dimensional computed tomography images of internal structures; 1.b Arrange the collected bioluminescent images into data vectors, and use the finite element method to construct the relationship between the data vectors and the distribution of unknown light sources in the body, as follows: y=Ax+n (1) In the formula, y is obtained from the bioluminescence image, and the size is M rows and 1 column, A is the coefficient matrix obtained from computed tomography images, the size of which is M rows and N columns, x is the unknown light source distribution in the small animal body, the size is N rows and 1 column, n is noise, the size is M rows and 1 column; 2. Set the initial value: Set the initial light source distribution x, the initial threshold β, and the initial sparsity Among them, x≥0, β≥1, 3. Iteratively update the light source distribution: 3.a Take the first row of the coefficient matrix, denoted as A ₁ , take the first element of the data vector Y, denoted as y ₁ , and calculate the increment: $\mathrm{<span\; class="notranslate">\; \&Delta;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\frac{{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span>}{{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$ In the formula, is the square of the l ₂ norm of A ₁ , γ is the weight, γ<1; 3.b Utilize the above increment △ to update x, and the updated x is denoted by x', which has: x'=x+Δ (3); 3.c Substitute x' obtained in step 3.b into formula (2), take the second row A ₂ of the coefficient matrix A and the second element y ₂ of the vector y, calculate the new increment Δ, and continue to use the formula ( 3) Update x; until the Mth row of the coefficient matrix A and the Mth element of the vector y are used to calculate the increment and update x is completed, the light source distribution after the update is recorded as _xnew ; 4. Threshold correction light source distribution: 4.a Take the largest element of x _new obtained in step 3.c, record it as x_max, and calculate $\mathrm{\&alpha;}=\frac{x\_\max }{\mathrm{\&beta;}};$ 4.b Compare the first element to the Nth element of x _new with α in step 4.a in turn, if the element is less than α, set the element to zero to obtain the corrected light source distribution, record for 4.c Use the following formula to calculate the above corrected light source distribution The sparsity of: $\mathrm{<span\; class="notranslate">\; \&psi;</span>}<span\; class="notranslate">\; =</span>\frac{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; new</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; /</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; new</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$ In the formula, respectively The l ₁ norm and l ₂ norm of ; 4.d Compare the sparsity ψ obtained in step 4.c with the initial sparsity To compare, if Then change the threshold β and return to steps 4.a, 4.b and 4.c for calculation and judgment; if Then the obtained threshold value and the corrected light source distribution are denoted as and Execute step five; ε is the error; 5. Calculation error and judgment stop condition: According to the corrected light source distribution in step 4.d calculate and will Compare with y in formula (1), if then get from step 4.d and As the initial threshold and initial light source distribution, ie and Return to step 3 and step 4, iteratively update the light source distribution and threshold value correction light source distribution again; if As the final light source distribution, denoted as x _opt , go to step 6; 6. Positioning of the light source: According to the final light source distribution x _opt obtained in step five, the maximum element position of the final light source distribution is found to complete the light source positioning. 2. The method for locating light sources for small animal bioluminescence imaging according to claim 1, characterized in that, in step 2, when setting the initial value, x=0, β=2, 3 . The method for locating light sources for small animal bioluminescent imaging according to claim 1 , wherein in step 3, γ=0.25 when iteratively updating the light source distribution to calculate the increment. 4 . The method for locating light sources for small animal bioluminescence imaging according to claim 1 , wherein, in steps 4 and 5, the error ε=1e ⁻⁶ . 5. The method for locating light sources for small animal bioluminescence imaging according to claim 1, characterized in that, when the quantitative optical molecular tomography device is used to acquire the bioluminescent images of the body surface of small animals and the computed tomography images of the internal structures, the posture of the small animals constant.