Dynamic Fluorescence Tomography Imaging Technology And Its Application In Pharmacokinetics-A Simulation Study

Fluorescence tomography, an important imaging modality, has been one of the hot spots in the field of molecular imaging. Although fluorescence tomography technology developed rapidly, the data frames of fluorescence tomography were independent,which didn’t take the correlation between the data and time into account. In this thesis we focused on the dynamic fluorescence tomography where time was correlated.Dynamic fluorescence tomography not only had advantages of the fluorescence tomography imaging, but also could describe the absorption completely, distribution and discharge of fluorophore in vivo, which was important for complex biological in drug research.How to reconstruct the three-dimensional distribution of the fluorophore was the key issue in fluorescence tomography technique. In this thesis, adaptive finite element method was used to reconstruct the distribution of fluorophore, and optimization methods were adopted in the process of fluorophore reconstruction. It was showed that hybrid regularization method got a better three-dimensional distribution of the fluorophore by comparison.Dynamic fluorescence tomography added time as a new variable dimension, so that the three-dimensional distribution of fluorophore based on the fluorescent yields of different fluorophore could be obtained, and which was used to analyze the time dynamic characteristics of fluorescence. Since the pharmacokinetic occurred in the organs with a definite volume, such as the heart, liver and kidneys, we used the L2 regularization to the initial mesh instead of the sparse regularization method and then apply the Landweber method to the subdivision mesh. Dynamic fluorescence tomography could reconstruct the concentration of fluorophore with different fluorescent yields; therefore, a curve of fluorophore concentration could be obtained.We could estimate the parameters based on the concentration curve, and gained the metabolic parameters. The nonlinear least squares method was used to build different compartment models, and subsequently we selected a suitable compartment model according to the weighted residuals and goodness of fit. Finally, we estimated pharmacokinetic equation coefficient to solve the drug metabolism rate parameters. The drug metabolism rate parameters showed the absorption and elimination of fluorophore in biological tissue, which was significant in the field of new drugs research.