北京理工大学 增强现实智能医疗实验室

Abstract—The visible human project has created an authoritive set of tomographic images of the human bodies and helped various medical applications. This paper reports a method to turn the grey-scale CT images of the visible human into dual-/multi-energy counterparts. An image dataset from dual-energy CT scanner was registered to the visible human CT images using a non-rigid registration method, and the waterand bone-basis maps were fused by the estimated deformation field onto the visible human CT images. Then, the monochromatic image dataset at any KeV can be obtained from the mixture of the two basis materials. Then, it becomes feasible to estimate the two basis materials from the visible human CT atlas, and synthesize a standard spectral visible human CT dataset.