基于VTK的医学图像可视化设计与实现  

The Design and Implementation of a Medical Image Visualization Based on VTK

作　　者：王华锋 李丹阳 

机构地区：[1]北方工业大学电子信息工程学院,北京100144

出　　处：《北方工业大学学报》2017年第5期8-12,共5页Journal of North China University of Technology

摘　　要：本文在VTK类库提供可视化与实现的基础上,首先运用卷积神经网路实现器官的识别与分割,然后分别采用移动立方体算法和光照投影法进行了器官图像的重建工作.本文提出识别与分割的方法达到了94.3%的准确率,重建工作可满足对医学图像与可视化的需要.Based on the visualization and implementation of VTK class library,this paper first uses the convolution neural network to realize the recognition and segmentation of organs,and then uses the mobile cube algorithm and the illumination projection method to reconstruct the images of organs respectively.With the proposed method of identification and segmentation in this paper the accuracy rate is high up to 94.3%,and the reconstruction work can meet the needs of medical image and visualization.

关 键 词：VTK 预处理 面绘制 体绘制 医学图像可视化 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]