Visible Human图象数据集的自动匹配方法

AUTOMATIC METHODS FOR REGISTRATION OF VISIBLE HUMAN DATASET

VisibleHuman计划是由美国国家医学图书馆发起建立的真实三维人体的CT、MRI和解剖切面的数字化图象库。VisibleHuman数据集已经成为构造电子医学图象库和虚拟解剖环境的理想基础。本文首先对CT图象进行灰度校正 ,并用基于像素相似性的方法实现了CT图象和MRI图象间的自动匹配。具有自适应空间搜索的遗传算法被用于实现全局性的优化过程。

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.