【作者】 张林海；

【导师】 冯先光； 杨武年；

【作者基本信息】 成都理工大学 ， 地图学与地理信息系统， 2008， 硕士

【摘要】 CT作为一种重要的医学图像扫描器件。目前已经成为放射诊断领域内不可缺少的一部分,是一种成熟的、在临床上普遍认可的检查方法。但是,任何技术的发展都有待提高和进一步改进的地方。CT是利用X射线来达到成像的目的,X射线摄影可以获得被摄影对象的平面影像。由于CT机用机械的方式模拟Radon变换的扫描无法一次确定每条X光线所经过的物体的体素,所以用准直系统屏蔽了大部分的X光线,这样使得每次扫描所获得的数据只能是物体上的一个断层或几个断层的数据。由此导致扫描所重建的图像只能是一个断面或几个断面,再由多个断面重建物体的容积图像。这种机械模拟的方式大大限制了其扫描速度,为了提高扫描速度CT技术采用的两种方式:增加探测器数量和改变扫描方式。这两种方式在增加CT制造难度的同时,没有取得更好的实际效果,重建出来的图像依然是断层图像。本文由CT应用的两大技术受到启发,应用了航空摄影测量学的定位方法提出了另一种CT扫描理论—X射线摄影重建容积图像。该理论借鉴了摄影测量的定位方法在被摄物体上做多个标志点,通过标志点坐标和其在影像上的坐标的转换关系,使得每次摄影可以同时确定入射到X像片平面的所有X射线光束路径。这样重建出来的图像是整个物体的容积图像。CV理论改变了传统CT扫描的定位方式,该理论的实现能提高扫描速度、降低CT机的制造难度、提高重建图像的分辨率等综合性能。同时,CV理论的实现改变了另一相关科学研究领域—医学图像体视化的方法和内容。基于CT断层图像的体视化需要将断层图像校正、插值等处理来通过不同的体视化方法显示物体的三维结构。而基于CV理论的体视化不需要这些操作,大大简化了体视化的方法步骤。为了更形象的说明基于这两种理论的体视化方法的异同,本文在MFC框架下调用OpenGL函数建立了一体视化框架,在这个框架下虚拟实现了基于CV理论的容积图像显示,并且对这一容积图像实现了两种不同的体视化显示方法,基于体视的显示方法和基于表面的显示方法。本文是一篇理论性的文章,所有的知识点都是围绕X射线摄影重建容积图像这一理论而展开的,通过相关知识点的融合、比较和论证,得出了一个结论,如果CV理论的能够得以实现,可以说相对于现有CT技术是一次大的变革。但是该理论的可行性还需要进一步的实践来加以检验。更多还原

【Abstract】 Computed Tomography is an important physical image scanning apparatus. In current days, CT has being becomes indispensable parts in radiology diagnosis fields and becomes maturated and indispensable inspect means in clinic. But, any technology always has itself disadvantage and need ameliorate place.CT makes use of X-ray to reconstruct image. X-ray photographing can get a flat image about scanned object. At present, the scanning methods of CT utilizes rigid machine slip ring to meet Radon transform that cause no way to fix on each X-ray pass through those object voxel. So using filtration system shield a majority of X-ray. That a scanning only get one or several fault scanning data, thereby CT can but to reconstruct one or several fault images. And then make using of mult-fault image to reconstruct volume image. The machine simulative scanning methods right smart restrict its scanning velocity, CT technology development mainly focus on how to heighten its scanning velocity and efficiency, for the sake of it adopt two means: increasing detector number and alter scanning means. Those means increase the difficultly of machine manufacturing moreover no attaining expected result.The paper obtains enlighten from CT application’s two technological principles and adopts Remote Sensing orientation theory to present the other one CT scanning theory-X-Ray Photographing Reconstruct Volumetric Image(CV). CV using photographing method put multi-marked points on photographed object.and depending on marked-points coordinate and its image coordinator’s transform relation. It can fix on X-ray path including all X-ray in X-image immediately. So the reconstructed image is three-dimensional volume image about scanned object. CV theory change orientation method about CT sanning image. Furthermore it comes true can debase the making difficulty of CT, heighten reconstruct image resolution and scanning velocity etc.Commonly, CV theory has altered the other one mutuality science research fields’ mehtods and contents about physical image volume visualization. Volume visualization based on fault image needs proofread and insert value fault image to display object’s three-dimensional structure by various volume visualization methods. And that volume visualization based on CV theory no needs the above manipulations and maximum reduce means and approaches. For more detailed account for volume visualization method’s distinguish based on the above two theory. This paper builded a frame of volume visualization adopting OpenGL API in MFC. And then realizing a volume visualization based on CV theory by virtual reality. Including realizing two vaious volume visualization display means surface based rendering and voxel based redering.This paper is an academic article, all knowledge focus on X-ray photographing reconstruct volumetric image. After syncretizing, comparing and demonstrating all viewpoints, as a result, that CV will become true, it is a new innovation for currently CT technology. But the theory feaibility is yet must more practicing and testing.更多还原