| 英文摘要 | Molecular imaging technique enables people to non-invasively reveal the molecular activities of living body in vivo. Compared with the traditional imaging modalities, molecular imaging can investigate the same objects continually, thus making the corresponding research more convincing. Currently, molecular imaging technique has been widely applied in the area of biomedicine and drug development. As an important part of molecular imaging, optical molecular imaging has the advantages of non-radiativity, high sensitivity, rapid measurement and low cost, etc. Therefore, it has been developing very fast in recent years. Fluorescence molecular tomography (FMT) is an important modality of optical molecular imaging. Different from the traditional two-dimensional imaging techniques, FMT can achieve three-dimensional imaging of molecular processes non-invasively, which makes it a hot spot in recent years. Although much progress has been reported, problems still remain in FMT. Because of the limitations of the current forward models and reconstruction methods, much effort is still needed to further improve the quality of the reconstruction results. In this thesis, researches on further improving the accuracy of the FMT reconstructions are conducted, and three new reconstruction methods, which are designed for different situations, are presented. These methods incorporate the sparsity of the fluorescent sources into the reconstruction, which is a very important a priori knowledge. Therefore, the quality of the results can become better. Besides the reconstruction methods, a new forward model that is based on the simplified spherical harmonics approximation is presented, which further improves the performance of FMT. The main contributions of this thesis are listed as follows: (1) We propose an iteratively reweighted reconstruction method for FMT. This method incorporates the sparse a priori information into the reconstructions. Sparsity is promoted by using sparsity-promoting regularization. In the reconstruction process, a weighting matrix is introduced, which is updated dynamically. When entering a new iteration, the weighting matrix is updated using the solution from the last iteration. By adopting this approach, we can use the reweighted L2-norm to approximate L0-norm and L1-norm. The reconstruction results show that this method can preserve the sparsity of the fluorescent sources and improve the quality of the image. (2) We propose a reconstruction method ... |
修改评论