Automatic methods for registration of visible human dataset

The Visible Human Project (VHP) is a digital image library of CT, MRI and cryosection images of the three-dimensional male and female human body developed by the National Library of Medicine. The Visible Human Project data sets tend to serve as a common reference point for the construction of medical image libraries and virtual anatomy environments. In this paper, CT images are correlated with gray level firstly, and methods based on voxel similarity are used to match the CT and MRI images automatically. Genetic algorithms with adaptive search space are used to establish the global optimization. Preliminary experiment results are given in this paper.