Bioluminescence tomography with structural information estimated via statistical mouse atlas registration.

摘要: Due to an ill-posed and underestimated characteristic of bioluminescence tomography (BLT) reconstruction, a priori anatomical information obtained from computed (CT) or magnetic resonance imaging (MRI), is usually incorporated improve the reconstruction accuracy. The organs need be segmented, which time-consuming challenging, especially for low-contrast CT images. In this paper, we present BLT method based on statistical mouse atlas efficiency heterogeneous model generation accuracy target localization. image was first registered with constraints surface high-contrast (bone lung). Then other organs, such as liver kidney, were determined automatically through model. estimated then discretized into tetrahedral meshes reconstruction. linearized Bregman used solve sparse inverse problem by minimizing regularization function (L1 norm plus L2 smooth factor). Both numerical simulations in vivo experiments conducted, results demonstrate that even though localization may not exactly accurate, proposed feasible reconstruct bioluminescent source effectively accurately organs. This would greatly benefit light hybrid BLT/CT systems.