A method, system, device and storage medium for rapid reconstruction of fluorescence molecular tomography

Abstract

The present invention belongs to the field of molecular imaging technology and discloses a method, system, device and storage medium for rapid reconstruction of fluorescence molecular tomography. The method excites a specific fluorescent probe through an excitation light source and uses a data acquisition system to obtain light distribution information on the surface of biological tissues. The method constructs a linear relationship between the surface light source distribution and the fluorescent target based on a light transmission model and a finite element method. The linear relationship obtained is converted into an optimization problem to be solved. The optimization problem is quickly solved using an adaptive fast iterative shrinkage threshold algorithm, and the results are displayed to complete the reconstruction. The present invention adopts multi-point excitation and multi-angle measurement, and at the same time combines the light transmission model and the finite element method to establish a linear relationship between the measurement data and the fluorescent target, and converts it into an optimization problem. The adaptive fast iterative shrinkage threshold algorithm is used to quickly solve the optimization problem to be solved, thereby ensuring the reconstruction accuracy while also speeding up the reconstruction speed, effectively improving the efficiency of fluorescence molecular tomography.

Description

Technical Field

[0001] The present invention belongs to the field of molecular imaging technology, and in particular relates to a method, system, device and storage medium for rapid reconstruction of fluorescence molecular tomography. Background Art

[0002] Fluorescence molecular tomography (FMT) is a new imaging modality developed in recent years with promising applications. It uses an external light source to excite specific fluorescent probes (fluorescent proteins, fluorescent dyes, etc.) to emit photons. A fluorescence acquisition device (a highly sensitive CCD camera) then collects the fluorescence signal from the surface of biological tissue. Combined with a light transmission model, the positional distribution and concentration of the fluorescent probes within the target can be reconstructed, enabling qualitative and quantitative studies of biological processes in vivo at the cellular and molecular levels. FMT is currently widely used in areas such as early disease diagnosis, efficacy monitoring, and new drug development.

[0003] The inverse problem of FMT is severely ill-posed. The fundamental reason is that the strong scattering properties of light cause photons to no longer propagate in straight lines within a biological organism, but instead undergo a large number of irregular scattering processes. Furthermore, because the acquired fluorescence distribution is confined to the surface of the imaging target and is limited in quantity, the inverse problem is a mathematical problem with a significantly greater number of unknowns than equations, further increasing the ill-posedness of the solution. To alleviate the ill-posedness of the inverse problem, increasing the number of projection data and multi-point excitation are often used, but this also incurs a significant time cost. Therefore, developing more suitable algorithms to improve reconstruction speed while maintaining accuracy is a key area of ​​focus for FMT.

[0004] Through the above analysis, the problems and defects of the existing technology are: in order to alleviate the pathological nature of the inverse problem, increasing projection data and multi-point excitation are often used as means, but this also brings huge time costs. Summary of the Invention

[0005] In response to the problems existing in the prior art, the present invention provides a method, system, device and storage medium for rapid reconstruction of fluorescence molecular tomography.

[0006] The present invention is implemented as follows: a method for rapid reconstruction of fluorescence molecular tomography, wherein the method utilizes a proposed adaptive rapid iterative shrinkage threshold algorithm to solve the acquired optical data. The algorithm is mainly improved by a new adaptive search strategy. The algorithm using this strategy has a faster convergence speed, thereby accelerating FMT reconstruction. Specifically, the method includes:

[0007] In the first step, a specific fluorescent probe is excited by an excitation light source and the light distribution information on the surface of biological tissue is obtained using a data acquisition system;

[0008] In the second step, the linear relationship between the surface light source distribution and the fluorescent target is constructed based on the light transmission model and the finite element method;

[0009] The third step is to transform the obtained linear relationship into the optimization problem to be solved;

[0010] In the fourth step, the adaptive fast iterative shrinkage threshold algorithm is used to quickly solve the optimization problem and the results are displayed to complete the reconstruction.

[0011] The first step is mainly to obtain information, which is a necessary part of the subsequent steps; the second step is mainly to establish a mathematical model for the data in step one; the third step is to convert the mathematical model into a mathematical problem that can be solved by subsequent algorithms; the fourth step is to propose a new algorithm to speed up the reconstruction.

[0012] Furthermore, the first step is implemented by using an excitation light source to excite a specific fluorescent probe in the organism, which emits an optical signal of a specific wavelength after being excited, and then using a data acquisition system to collect the optical signal to obtain light distribution information on the surface of the biological tissue.

[0013] Furthermore, an excitation light source is used to excite specific fluorescent probes in the body, and multi-point excitation and multi-angle measurement are used to obtain large-scale fluorescence data.

[0014] Furthermore, fluorescence measurement data and the anatomical structure information and optical characteristic parameters of the reconstructed object are obtained. The imaging target is discretized using Amira software to obtain a finite element mesh. The implementation process is as follows: Based on the diffusion approximation model and Robin boundary conditions, the system matrix A is used to establish a linear relationship between the fluorescence information Φ obtained by surface measurement and the fluorescent target to be reconstructed X:

[0015] AX = Φ;

[0016] Where X represents the fluorescent target to be reconstructed; Φ represents the fluorescence distribution information measured on the surface; and A is the system matrix.

[0017] Furthermore, the third step is specifically to transform the above linear relationship into a convex optimization problem. The distribution of fluorescent molecular probes in biological tissues is relatively sparse, so the linear relationship described in claim 4 can be transformed into an l1 norm minimization problem:

[0018]

[0019] λ is the regularization parameter.

[0020] Furthermore, the fourth step uses an adaptive fast iterative shrinkage threshold algorithm to accelerate the solution of the model. Write it as f(X), for a continuously differentiable function f, the following Lipschitz continuity condition is satisfied:

[0021]

[0022] Where L(f) is the Lipschitz constant. By introducing the proximal operator, for any point Y, the minimization problem is transformed into the following problem:

[0023]

[0024] Q L There is a unique minimum point in (X, Y):

[0025]

[0026] According to the iterative shrinkage threshold algorithm, the following iterative steps are obtained:

[0027] x k =p L (x k-1 );

[0028] The Lipschitz constant determines the step size of the iteration. During each iteration, the proximal gradient condition should be satisfied:

[0029]

[0030] An adaptive search strategy is introduced to improve the accuracy of the Lipschitz constant, thereby accelerating the algorithm. The adaptive search strategy includes the following steps:

[0031] Step 1: Construct the amplification factor η based on the gradient of each iteration k , when the proximal gradient condition is not met, the Lipschitz constant is amplified as follows:

[0032] L k =(1+η k )*L k

[0033] d k is the gradient operator, and

[0034]

[0035] α′ is such that L k The maximum value when the proximal gradient condition is met is updated by α′=0.9*α′ in each backtracking.

[0036] Step 2: Determine whether the Lipschitz constant calculated between the two iterations has changed. If not, reduce the Lipschitz constant using the following formula:

[0037] L k =p k *L k-1 ;

[0038] The reduction factor p k Defined as:

[0039]

[0040] where η k is the amplification factor obtained in step 1;

[0041] The fast iterative shrinkage threshold algorithm uses the Nesterov strategy to speed up the convergence speed, that is, to include t k+1 With t k The auxiliary sequence is realized with x k with x k-1 The next iteration point is obtained by a special linear combination of ; after the introduction of the adaptive search strategy, the value of the Lipschitz constant selected in this iteration may be smaller than the value of the Lipschitz constant selected in the previous step. The original auxiliary sequence cannot guarantee the convergence of the algorithm. In order to ensure the convergence of the algorithm, the following inequality is satisfied:

[0042]

[0043] where v k =F(x k )-F(x * ),u k =t k x k -(t k -1)x k-1 -x * , x * is the optimal value point;

[0044] Construct a new auxiliary sequence After introducing the new sequence, the iterative steps of the algorithm are:

[0045]

[0046]

[0047]

[0048] Furthermore, the fourth step specifically includes:

[0049] (4a) Initialize the parameters: Lipschitz constant L0, expansion factor η0 and reduction factor p0, initial iteration point x0;

[0050] (4b) Enter the adaptive fast iterative shrinkage threshold algorithm and iteratively solve until the iteration condition is met and the output result X is output;

[0051] (4c) Use Tecplot software to construct the three-dimensional image results, complete the reconstruction and calculate various indicators.

[0052] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the fluorescence molecular tomography rapid reconstruction method.

[0053] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the fluorescence molecular tomography rapid reconstruction method.

[0054] Another object of the present invention is to provide a fluorescence molecular tomography rapid reconstruction system for implementing the fluorescence molecular tomography rapid reconstruction method, wherein the fluorescence molecular tomography rapid reconstruction system comprises:

[0055] A distribution information acquisition module is used to excite a specific fluorescent probe through an excitation light source and obtain light distribution information on the surface of biological tissue using a data acquisition system;

[0056] A linear relationship building module is used to build a linear relationship between the surface light source distribution and the fluorescent target based on the light transmission model and the finite element method;

[0057] A problem-solving module, used to convert the obtained linear relationship into an optimization problem to be solved;

[0058] The imaging construction module is used to quickly solve the optimization problem using an adaptive fast iterative shrinkage threshold algorithm and display the results to complete the reconstruction.

[0059] In combination with the above technical solutions and the technical problems solved, please analyze the advantages and positive effects of the technical solutions to be protected by the present invention from the following aspects:

[0060] First, in view of the technical problems existing in the above-mentioned prior art and the difficulty of solving these problems, this paper closely combines the technical solutions to be protected by the present invention and the results and data during the research and development process, and analyzes in detail and in depth how the technical solutions of the present invention solve the technical problems and some creative technical effects brought about by solving the problems. The specific description is as follows:

[0061] Multi-point excitation and multi-angle measurement are adopted, and the linear relationship between the measurement data and the fluorescent target is established by combining the light transmission model and the finite element method, and it is converted into an optimization problem. For the optimization problem to be solved, an adaptive fast iterative shrinkage threshold algorithm is used for rapid solution. The standard fast iterative shrinkage threshold algorithm adopts a linear growth search strategy. If the algorithm calculates a large value of the Lipschitz constant at the beginning of the iteration, the algorithm will always maintain a small step size, thereby affecting the convergence speed of the algorithm. The algorithm of the present invention introduces an adaptive search strategy that combines gradient information to improve the accuracy of the step search. A more accurate step search will effectively improve the convergence speed of the algorithm. Therefore, using this method for FMT reconstruction can not only ensure the reconstruction accuracy, but also speed up the reconstruction speed. The present invention effectively improves the efficiency of fluorescence molecular tomography and has important application value in the fields of medical molecular imaging, reconstruction methods, etc.

[0062] Second, considering the technical solution as a whole or from the perspective of the product, the technical effects and advantages of the technical solution to be protected by the present invention are described in detail as follows:

[0063] This method uses multi-point excitation and multi-angle measurement to obtain large-scale surface light data. Simultaneously, based on the light transmission model and the finite element method, a linear relationship between the surface light data and the distribution of the fluorescent probes to be reconstructed is established. During the reconstruction process, the proposed adaptive rapid iterative shrinkage threshold algorithm is used to solve the inverse problem, which not only ensures reconstruction accuracy but also improves the reconstruction speed. Overall, the present invention is based on hardware equipment and centered on software algorithms, and is a complete solution for achieving rapid FMT reconstruction.

[0064] Third, as auxiliary evidence for the inventiveness of the claims of the present invention, it is also reflected in the following important aspects:

[0065] (1) The expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0066] The reconstruction method described in this invention is used to accelerate FMT reconstruction and can be extended to new imaging modalities such as bioluminescent tomography and X-ray luminescence tomography. FMT is widely used in areas such as early disease diagnosis, efficacy monitoring, and new drug development, and is expected to be in high demand in the future medical market and have significant commercial value. This method can effectively improve the rapid reconstruction of FMT and reduce the time cost of FMT applications, undoubtedly providing a reliable method that will promote the development of FMT.

[0067] (2) The technical solution of the present invention fills the technical gap in the industry at home and abroad:

[0068] The reconstruction method described in the present invention is the first in China and abroad to introduce an adaptive search strategy that combines gradient information. The proposed adaptive search strategy is more flexible than the linear search strategy of the fast iterative shrinkage threshold algorithm and can obtain a better step size, thereby improving the reconstruction speed of the fast iterative shrinkage threshold algorithm in FMT. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a flow chart of a rapid reconstruction method for fluorescence molecular tomography provided by an embodiment of the present invention;

[0070] Figure 2 Schematic diagram of the structure of a rapid reconstruction system for fluorescence molecular tomography provided by an embodiment of the present invention;

[0071] Figure 3 is a flowchart of the implementation of the fluorescence molecular tomography reconstruction method provided by an embodiment of the present invention;

[0072] Figure 4 This is a calculation flow chart of the adaptive fast iterative shrinkage threshold algorithm provided by an embodiment of the present invention;

[0073] Figure 5 Schematic diagram of a digital mouse model for simulation experiments provided by an embodiment of the present invention;

[0074] Figure 6 1 is a reconstruction result obtained by the reconstruction algorithm provided in an embodiment of the present invention, wherein (a) is a 3D reconstruction result, and (b) is a 2D result diagram of a section Z=16.0 mm;

[0075] Figure 7 1 is a parameter comparison diagram of the reconstruction algorithm (AFISTA) and the fast iterative shrinkage threshold algorithm (FISTA) provided by an embodiment of the present invention, wherein Figure (a) compares the iteration curves of the two algorithms, and Figure (b) compares the quantitative results of the two algorithms;

[0076] In the figure: 1. Distribution information acquisition module; 2. Linear relationship construction module; 3. Problem solving module; 4. Imaging construction module. DETAILED DESCRIPTION

[0077] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0078] 1. Explanatory Examples In order to enable those skilled in the art to fully understand how to implement the present invention, this section provides an illustrative example that expands upon the technical solutions of the claims.

[0079] like Figure 1As shown, the fluorescence molecular tomography rapid reconstruction method provided by the embodiment of the present invention includes the following steps:

[0080] S101: Exciting a specific fluorescent probe through an excitation light source and obtaining light distribution information on the surface of biological tissue using a data acquisition system;

[0081] S102: Constructing the linear relationship between surface light source distribution and fluorescent target based on light transmission model and finite element method;

[0082] S103: converting the obtained linear relationship into an optimization problem to be solved;

[0083] S104: Using an adaptive fast iterative shrinkage threshold algorithm to quickly solve the optimization problem, display the results and complete the reconstruction.

[0084] like Figure 2 As shown, the fluorescence molecular tomography rapid reconstruction system provided by the embodiment of the present invention includes:

[0085] The distribution information acquisition module 1 is used to excite the specific fluorescent probe through the excitation light source and obtain the light distribution information on the surface of the biological tissue using the data acquisition system;

[0086] Linear relationship construction module 2, used to construct the linear relationship between the surface light source distribution and the fluorescent target based on the light transmission model and the finite element method;

[0087] Problem solving module 3, used for converting the obtained linear relationship into an optimization problem to be solved;

[0088] The imaging construction module 4 is used to quickly solve the optimization problem using an adaptive fast iterative shrinkage threshold algorithm, and display the results to complete the reconstruction.

[0089] like Figure 3 As shown, the fluorescence tomography reconstruction method provided by the embodiment of the present invention specifically includes the following steps:

[0090] (1) Excite the fluorescent probe in the organism and use the data acquisition system to obtain the light distribution information on the surface of the biological tissue.

[0091] (1a) Multi-point excitation is used to acquire large-scale data. The specific fluorescent probe can be cy5.5, ICG and other materials.

[0092] (1b) The acquisition system is an imaging system with a CCD camera, such as a multimodal imaging system with fluorescence acquisition and CT capabilities. Using projection tomography, a laser and an optical detection instrument are placed on either side of the imaging target. The laser illuminates the reconstruction target, stimulating the fluorescent target to emit fluorescence. The fluorescence penetrates the imaging target and is detected by the optical detector on the opposite side.

[0093] (2) Constructing the linear relationship of surface light source distribution based on the light transmission model and finite element method;

[0094] Finite element meshing of the imaging target organism is performed using Amira software. Then, the diffusion approximation equation is discretized according to the finite element theory to construct the linear equation for the surface measurement data and the distribution of the fluorescent target inside the reconstructed target:

[0095] AX = Φ;

[0096] Where X represents the fluorescent target to be reconstructed; Φ represents the fluorescence distribution information measured on the surface; and A is the system matrix.

[0097] (3) Convert the linear relationship described in (2) into a 1-norm minimization problem. The target to be reconstructed can be considered sparse in the entire area, so the linear relationship can be converted into a 1-norm minimization problem:

[0098]

[0099] (4) Figure 4 As shown, the model obtained from (3) is quickly solved using an adaptive fast iterative shrinkage threshold algorithm.

[0100] (4a) For the problem to be solved, the minimization problem in (3) can be Write it as f(X), then for a continuously differentiable function f, the following Lipschitz continuity condition is satisfied:

[0101]

[0102] Where is the Lipschitz constant. By introducing the proximal operator, for any point Y, the optimization problem in (3) is transformed into the following problem:

[0103]

[0104] Q L There is a unique minimum point in (X, Y):

[0105]

[0106] After the introduction of the proximal operator, the following iterative steps can be obtained according to the iterative shrinkage threshold algorithm:

[0107] x k =p L (x k-1 );

[0108] (4b) A search strategy to accelerate convergence is introduced, which is to adaptively expand or reduce the Lipschitz constant according to the gradient information, where the expansion factor is:

[0109]

[0110] d k is a gradient operator, and:

[0111]

[0112] α′ is such that L k The maximum value when the proximal gradient condition is met is updated by α′=0.9*α′ in each backtracking.

[0113] The Lipschitz constant L is amplified using the amplification factor:

[0114] L k =(1+η k )*L k ;

[0115] When the value of the Lipschitz constant L remains unchanged between the first and last iterations, it is reduced:

[0116] L k =p k *L k-1 ;

[0117] The reduction factor is defined as:

[0118]

[0119] where η k is the amplification factor obtained in step 1.

[0120] (4c) Introducing auxiliary sequences Its subsequent iterations are:

[0121]

[0122]

[0123]

[0124] (5) Display the results. The results obtained in step (4) are processed using Tecplot software and then displayed.

[0125] 2. Application Examples: In order to demonstrate the creativity and technical value of the technical solution of the present invention, this section provides application examples of the claimed technical solution on specific products or related technologies.

[0126] The target on the rotating stage is excited by transmission using an excitation light source. In this embodiment, the rotating stage is controlled by a computer to rotate at equal intervals, with excitation occurring every 60 degrees. A CCD camera is used for data acquisition, thereby achieving multi-angle excitation and simultaneously obtaining multiple sets of measurement data. After obtaining these multiple sets of measurement data, a linear relationship between the measured fluorescence information on the surface and the target to be reconstructed is constructed. An adaptive fast iterative shrinkage threshold algorithm is used for rapid solution, and the obtained results are processed and displayed using Tecplot software.

[0127] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0128] 3. Evidence of the effects of the embodiments: The embodiments of the present invention have achieved some positive effects during the development or use process, and indeed have great advantages over the existing technology. The following content describes them with reference to the data, charts, etc. of the experimental process.

[0129] Figure 5 This is a digital mouse model used in simulation experiments. The mouse's trunk was selected and its major organs, including the heart, lungs, liver, stomach, kidneys, and muscle tissue, were segmented.

[0130] Figure 6 This is the reconstruction result based on the present invention. Figure (a) is the 3D reconstruction result, and Figure (b) is the cross-sectional view at z = 16 mm. The center positions are (12.0, 8.0, 16.0) mm, and the reconstructed fluorophore center position obtained by the algorithm is (12.11, 8.00, 15.60) mm. The positioning error is defined as: The reconstruction time is 47.20s, and the number of iterations required for the reconstruction to converge is 175.

[0131] Figure 7This is a comparison of the results of the reconstruction algorithm of the present invention and the standard fast iterative threshold shrinkage algorithm (FISTA). Figure (a) is a convergence curve diagram of the two, and Figure (b) is the quantitative results of various parameters. As can be seen from Figure (b), the positioning error of the reconstruction algorithm of the present invention is slightly better than that of FISTA, and its positioning accuracy is improved by about 0.02mm. Based on the reconstruction algorithm of the present invention, while ensuring accuracy, the number of iterations is reduced by about 85% and the speed is increased by about 69%. It is an effective reconstruction method for fluorescence molecular tomography. During the reconstruction process, the Lipschitz constant of the standard fast iterative threshold shrinkage algorithm always maintains a large value, resulting in slow convergence of the algorithm. The value of the Lipschitz constant of the reconstruction algorithm of the present invention is adjusted according to the gradient information, and a constant value closer to the true value is obtained during iteration, which improves the convergence speed of the algorithm.

[0132] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A method for rapid reconstruction of fluorescence molecular tomography, characterized in that: The fluorescence molecular tomography rapid reconstruction method uses an adaptive rapid iterative shrinkage threshold algorithm to solve the collected optical data; The fluorescence molecular tomography rapid reconstruction method comprises: In the first step, a specific fluorescent probe is excited by an excitation light source and the light distribution information on the surface of biological tissue is obtained using a data acquisition system; In the second step, the linear relationship between the surface light source distribution and the fluorescent target is constructed based on the light transmission model and the finite element method; The third step is to transform the obtained linear relationship into the optimization problem to be solved; In the fourth step, the optimization problem is quickly solved using an adaptive fast iterative shrinkage threshold algorithm, and the results are displayed to complete the reconstruction; The fourth step uses an adaptive fast iterative shrinkage threshold algorithm to accelerate the solution of the model. Write it as f(X), for a continuously differentiable function f, the following Lipschitz continuity condition is satisfied: Where X represents the fluorescent target to be reconstructed; Φ represents the fluorescence distribution information measured on the surface; A is the system matrix, and L(f) is the Lipschitz constant. By introducing the proximal operator, for any point Y, the minimization problem is transformed into the following problem: Q L There is a unique minimum point in (X, Y): According to the iterative shrinkage threshold algorithm, the following iterative steps are obtained: x k =p L (x k-1 ); The Lipschitz constant determines the step size of the iteration. During each iteration, the proximal gradient condition should be satisfied: An adaptive search strategy is introduced to improve the accuracy of the Lipschitz constant, thereby accelerating the algorithm. The adaptive search strategy includes the following steps: Step 1: Construct the amplification factor η based on the gradient of each iteration k , when the proximal gradient condition is not met, the Lipschitz constant is amplified as follows: d k is the gradient operator, and α′ is such that L k The maximum value when the proximal gradient condition is met, that is, it is updated with α′=0.9*α′ in each backtracking; Step 2: Determine whether the Lipschitz constant calculated between the two iterations has changed. If not, reduce the Lipschitz constant using the following formula: L k =p k *L k-1 ; The reduction factor p k Defined as: where η k is the amplification factor obtained in step 1; The fast iterative shrinkage threshold algorithm uses the Nesterov strategy to speed up the convergence speed, that is, to include t k+1 With t k The auxiliary sequence is realized with x k with x k-1 The special linear combination of is used to obtain the next iteration point; after the introduction of the adaptive search strategy, the value of the Lipschitz constant obtained in this iteration may be smaller than the value of the Lipschitz constant selected in the previous step. The original auxiliary sequence cannot guarantee the convergence of the algorithm. In order to ensure the convergence of the algorithm, the following inequality is satisfied: where v k =F(x k )-F(x * ),u k =t k x k -(t k -1)x k-1 -x * , x * is the optimal value point; Construct a new auxiliary sequence After introducing the new sequence, the iterative steps of the algorithm are:

2. The method for rapid reconstruction of fluorescence molecular tomography according to claim 1, wherein: The first step is to use an excitation light source to excite a specific fluorescent probe in the organism, so that the fluorescent probe emits an optical signal of a specific wavelength after being excited, and then use a data acquisition system to collect the optical signal to obtain light distribution information on the surface of the biological tissue; The excitation light source is used to excite specific fluorescent probes in the body, and multi-point excitation and multi-angle measurement are used to obtain large-scale fluorescence data.

3. The method for rapid reconstruction of fluorescence molecular tomography according to claim 1, wherein: The fluorescence measurement data and the anatomical structure information and optical characteristic parameters of the reconstructed object are obtained. The finite element mesh is obtained by discretizing the imaging target using Amira software. The implementation process is as follows: Based on the diffusion approximation model and Robin boundary conditions, the system matrix A is used to establish the linear relationship between the fluorescence information Φ obtained by surface measurement and the fluorescent target to be reconstructed X: AX = Φ; Where X represents the fluorescent target to be reconstructed; Φ represents the fluorescence distribution information measured on the surface; and A is the system matrix.

4. The rapid reconstruction method for fluorescence molecular tomography according to claim 1, wherein: The third step is to transform the above linear relationship into a convex optimization problem. Since the distribution of fluorescent molecular probes in biological tissues is relatively sparse, the linear relationship can be transformed into an l1 norm minimization problem: λ is the regularization parameter.

5. The rapid reconstruction method for fluorescence molecular tomography according to claim 1, wherein: The fourth step specifically includes: (4a) Initialize the parameters: Lipschitz constant L0, expansion factor η0 and reduction factor p0, initial iteration point x0; (4b) Enter the adaptive fast iterative shrinkage threshold algorithm and iteratively solve until the iteration condition is met and the output result X is output; (4c) Use Tecplot software to construct the three-dimensional image results, complete the reconstruction and calculate various indicators.

6. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the fluorescence molecular tomography rapid reconstruction method according to any one of claims 1 to 5.

7. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor is caused to execute the steps of the fluorescence molecular tomography rapid reconstruction method according to any one of claims 1 to 5.

8. A fluorescence molecular tomography rapid reconstruction system for implementing the fluorescence molecular tomography rapid reconstruction method according to any one of claims 1 to 5, characterized in that: The fluorescence molecular tomography rapid reconstruction system comprises: A distribution information acquisition module is used to excite a specific fluorescent probe through an excitation light source and obtain light distribution information on the surface of biological tissue using a data acquisition system; A linear relationship building module is used to build a linear relationship between the surface light source distribution and the fluorescent target based on the light transmission model and the finite element method; A problem-solving module, used to convert the obtained linear relationship into an optimization problem to be solved; The imaging construction module is used to quickly solve the optimization problem using an adaptive fast iterative shrinkage threshold algorithm and display the results to complete the reconstruction.

Images