中文摘要：

为了利用CT、核磁共振成像（MRI）等医学图像确定病变部位的大小、形状与周围组织的空间关系,本文提出一套完整的数据源获取、点云压缩、旋转和显示的方案。以CT、MRI医学图像为数据源,运用Mimics软件对人体组织MRI图像进行分割以及三维重建。针对数据量的大问题,提出基于八叉树均匀化的压缩算法;为了医生多视角观察模型周围情况,提出基于柱坐标系的旋转算法,在点云库中实现上述算法。最后模型在体扫描显示屏上显示,辅助医生准确确定病变部位的情况,为手术规划提供一种模拟平台。

英文摘要：

A complete set of data source acquisition, point cloud compression, rotation and display program was proposed in the paper to accurately determine the size, shape of pathological positions and their spatial relationship with surrounding tissues by observing CT and MRI images. With the data source of CT and MRI images, human tissue MRI images were segmented and reconstructed by using Mimics software. Regarding the large amount of cloud data, the compression algorithm based on octree homogenization was proposed, and the rotation algorithm based on the cylindrical coordinate system was proposed to assist the doctor in observing the surrounding condition of model from multi perspectives. The proposed algorithms were implemented in point cloud library. Finally, the model was displayed on a three-dimensional swept volume display screen, assisting doctors in accurately determining the pathological location, and providing a simulation platform for surgical planning.