基于GPU的医学图像多功能可视化的实现

摘要

三维可视化在医学诊断过程中发挥着越来越重要的作用,随着GPU的高速发展,成像质量高的光线投射算法克服了其耗时的难题,越来越多地被使用到实际运用中.为了更好地提供用户体验,为医生诊断提供更多的服务,提出一种交互性多功能的医学图像三维可视化.以Visual Studio 2008为平台,DCMTK读取DICOM图像,CUDA (Compute Unified Device Architecture)进行光线投射算法的加速,实现了三维图像中定位,任意断面显示,区域的切割与特殊显示功能.实验结果显示该系统在普通台式电脑上具有较好的实时性和交互性,可以为医生提供更好的服务,拥有巨大应用价值.%Three-dimensional visualisation is playing a more and more important role in medical diagnosis.With the rapid development of GPU,the raycasting algorithm,which has high image quality,has overcome the problem of its time-consuming and has been increasingly used to practical application.In order to provide a better user experience and more diagnostic services for the physicians,we propose a three-dimensional visualisation of medical images with interactive multi-function.We use Visual Studio 2008 as the platform,use DCMTK to read DICOM images,and CUDA to accelerate ray casting algorithm,achieve the functions of localisation,arbitrary cross-section display,regional cutting and special display in three-dimensional images.Experimental results show that on typical desktop PC,the system with pretty good real-time property and interactivity can give better service to doctors and has great application value.