基于GPU的医学图像三维重建体绘制技术综述

摘要

Volume rendering technology is good for the display of objects' inside details, and has good practical value for medical diagnosis. But the currently used for the reconstruction of the image data increased significantly compared with previous, makes the traditional volume rendering based on CPU in 3D reconstruction of the drawing speed is slow and the interaction is not smooth. Thus volume rendering technology based on Graphic Processing Unit(GPU)has great research value, and with the rapid development of the parallel computing ability and the programming, the medical image real-time rendering based on GPU is possible. In this paper, firstly, we will introduce briefly the volume rendering technology. Secondly, we will state systematically the GPU. Then, we will demonstrate implementation of volume rendering technology based on GPU. In the end, we will summary and look into the future of volume rendering technology based on GPU.%体绘制技术能够很好地显示对象内部的细节信息，对于医学诊断有着很好的实用价值。但由于当前用于重建的图像数据量与以往相比有显著增加，传统的基于CPU的体绘制技术在三维重建中存在着绘制速度慢、交互不流畅等问题，从而使基于图形处理芯片（GPU）的体绘制技术有了很大研究价值。随着GPU并行计算能力和可编程性的快速发展，GPU用于医学图像实时绘制成为可能。首先对体绘制技术进行简单介绍，并对GPU进行系统阐述，然后对一种基于GPU的体绘制技术的实现做出综述，最后对基于GPU的体绘制技术进行总结和展望。