Research On Three-dimensional Reconstruction Method And Signal Enhancement Technology Of Radioactive Source In Cerenkov Fluorescence Imaging Imaging

Cerenkov luminescence imaging?CLI?technology is an optical molecular imaging using visible near-infrared light produced in the process of radionuclide decay.Because a large number of radionuclides can be used in clinical,CLI provides a new idea to solve the limitations of molecular probes in the clinical conversion of optical molecular imaging technology.Combining the light transmission model and light source reconstruction algorithm in biological tissue,Cerenkov fluorescence three-dimensional imaging?CLT?can accurately obtain the in vivo position and quantitative spatial distribution information of nuclide probe,so it is of great concern.However,due to the wide spectral characteristics of Cherenkov fluorescence,the inadequacy of Cherenkov fluorescence measurement data,and the strong decay of Cherenkov fluorescence of deep nuclide probes,CLT technology still faces great challenges in accurately obtaining the three-dimensional position and quantitative spatial distribution of nuclide probes in vivo.In this paper,Cherenkov fluorescence imaging technology are discussed and studied in three-dimensional reconstruction accuracy and deep target imaging effectiveness.The main work is summarized as follows1.Aiming at the inaccuracy of imaging model caused by the wide spectral characteristics of Cherenkov fluorescence,a multispectral Cherenkov fluorescence3D imaging method based on hybrid light transmission model is proposed.Using wide spectrum Cherenkov fluorescence for multispectral 3D imaging can solve the problem of incomplete measurement data to a certain extent,but it also causes the inaccuracy of the existing optical transmission model based on a single equation.Firstly,an automatic hybrid light transfer model is constructed to describe the transfer process of multispectral Cherenkov fluorescence in biological tissues.The hybrid equation is used in different spectrum bands and different tissues in the same spectrum bands,so the model has higher accuracy,efficiency and flexibility.Secondly,the finite element method and the multispectral data coupling strategy are used to discretize the hybrid model,and a linear system equation is established to describe the relationship between the internal target and the external measurement value.Finally,considering the sparsity of light source and the insufficiency of the measured value,the sparse regularization objective function based on l₁ norm is established,and the appropriate optimization algorithm is selected to solve the problem,to obtain the spatial position and quantitative distribution information of nuclide probe in vivo.The feasibility and validity of this method are verified by the simulation experiment of digital mouse and the experiment of real mouse,as well as the superiority in balancing the imaging efficiency and accuracy.2.Aiming at the problem that the reconstruction result tends to the surface from the single view measurement data,a Cerenkov fluorescence three-dimensional imaging based on a prior compensation is proposed.Single view measurement data collection is the most consistent way the real experimental data collection of commercial system,which can meet the limitation of time efficiency and collection conditions in practical application;however,single view measurement will bring the lack of data,which leads to the uncertainty of reconstruction results tend to the surface.Firstly,considering the influence of the depth of the source caused by the lack of information on the surface measurement signal,the depth effect compensation matrix is constructed to correct the influence of the depth change on the system matrix,and the system equation of fusion depth compensation is established.Secondly,considering the sparsity of target distribution in vivo,the l_p norm based sparse regularization strategy is used to transform the system equation into a regularization problem,and finally the spatial location and intensity distribution information of target in different depth in vivo is obtained.Through a series of simulation experiments in different depths,the accuracy and effectiveness of the method are proved,and the advantages is proved compared with the non compensation method.Experiments with live mice have shown that this method has better application prospects in pre-clinical research.3.Aiming at the imaging effectiveness caused by the strong attenuation of Cherenkov fluorescence of deep nuclide probe,an endoscopic Cerenkov fluorescence three-dimensional imaging based on the hybrid light transmission model is proposed.Endoscopic data acquisition provides a new way to solve the problem of limited imaging depth in clinical application of CLI technology.First of all,according to the characteristics of endoscopic data acquisition,a general imaging physical model is constructed,which is coupled with the boundary conditions of each part and the conversion of physical quantities,and a hybrid optical transmission mathematical model based on simplified spherical harmonic approximation,diffusion approximation and radiometry theory?sp3-da-radiosity?is established.Secondly,by combining the finite element discrete framework,the system equation between the internal target and the measurement value of the endoscope instrument is established.Finally,considering the sparsity of target in vivo and the serious deficiencies of endoscopic measurement value,the sparse regularization objective function based on l₁norm is established,and solved using the initial dual interior point method.The simulation results based on heterogeneous model verify the accuracy and effectiveness of the method.4.Aiming at the problem of imaging effectiveness caused by the strong attenuation of Cherenkov fluorescence of deep nuclide probe,a new type of Cherenkov fluorescence enhanced imaging technology is proposed,named radiofluorescence film imaging technology.Cherenkov fluorescence is the secondary product of radionuclide decay,which has the defects of weak signal and limited imaging depth,and seriously limits its clinical application.Inspired by the phenomenon of radiation fluorescence,that is,rare earth nanoparticles can emit fluorescence when bombarded with high-energy rays.This chapter proposes a imaging technology of radiation fluorescence film to detect radionuclides.This technology uses the film made of the radioactive fluorescence particles to stick on the surface of the imaging body,and carries out optical imaging based on the radiation fluorescence signal excited by the high-energy ray.In this way,compared with the Cherenkov fluorescence imaging technology,the radioactive fluorescence film imaging technology can detect the medical isotopes in the deeper position of the living object.Through a series of feasibility verification experiments,performance characterization experiments and in vivo mouse experiments,it has been proved that it has strong preclinical and clinical application potential.