Time-Domain Fluorescence Diffusion Optical Tomography-Fundamental Theory And Experiment Validation

As near infrared light can travel several centimeters in tissue, fluorescence diffuse optical tomography(FDOT)with the aid of specific fluorescent probes promises to open new pathways for the characterization of biological processes in living animals at cellular and molecular levels. The time domain technique offers the potential advantages of directly measuring lifetime and has the favorite performances of simultaneously recovering of fluorescent yield and lifetime distributions, as well as resolving multiple components. The thesis focuses on the research of the imaging theory and method on time-domain FDOT, including measurement method based on the multichannel time-resolved system, the algorithm of image reconstruction, numerical simulation and experimental investigation.The diffusion equation (DE) is the P1 approximation of the Radiative Transfer Equation, and has been proved to be adequately accurate in case of thick tissue and mathematically tractable. The fluorescent light detection relies on a coupled DE that associates the excitation-light with the emission-one. We propose a generalized pulse spectrum technique (GPST) for time-domain FDOT. In this work, a finite element method solution to the Laplace transformed time-domain coupled diffusion equations is employed as the forward model, and the resultant linear inversions at two distinct transform-factors are solved with an algebraic reconstruction technique to separate fluorescent yield and lifetime images. The normalized Born ratio is used for its independence of the source intensity and less sensitivity to the systematic errors. In addition, it eliminates the requirement for accurate calibration of the temporal-origin in time-domain measurement and also exhibits a high robustness to the uncertainties of highly optically heterogeneous background. The algorithm is validated using simulated data, and the spatial resolution, noise-robustness and so on are assessed. The choice of appropriate transform factors is discussed.The full time-resolved mode possesses the richest information about the optical and fluorescent properties. Making full use of the time-resolved data would improve fidelity of the image reconstruction, and help set up the'golden standard'for evaluating the performance of the other featured-data methodologies. In this work, a hybrid finite-element-finite-time-difference method for solving the TD coupled diffusion equations is developed to serve as the forward model. The Newtown-Raphson inverse model for simultaneous reconstruction of fluorescent yield and lifetime images is proposed, and some issues associated to its implementation, such as the computation of the weighting matrix, linear inversion and regularization strategy, are highlighted.The proposed algorithm is validated with simulated data for 2D phantom and its superiority in the improvement of image quality is demonstrated, in comparasion with the featured-data algorithm previously developed based on GPST.A method for determining the optical properties in turbid medium is developed based on time-resolved reflection measurement, and the optical properties of solid and liquid phantoms are measured. By use of multchannels time-resolved measurent system based on time-correlated single photon counting (TCSPC), we experimentally validate that the proposed GPST scheme can achieve simultaneous reconstruction of the fluorescent yield and lifetime distributions with a reasonable accuracy. The results demonstrate that the proposed methodology is suitable for further application into FDOT. Nevertheless, a lot of improvements on both the instrumentation and methodology are necessarily required prior to clinical applications.