【作者】 李侃；

【导师】 肖国臻；

【作者基本信息】 浙江大学 ， 计算机应用技术， 2006， 硕士

【摘要】 PACS(图像归档与通信系统)作为医院数字化的信息平台已经变的越来越重要,当前很多医院都在部署实施PACS系统。针对不同的医院规模和需求,PACS可以分为全院、科室、单机三种应用方式。PACS往往都包括信息登记、看图、中央调度以及实现其他子功能的模块。通过PACS平台,医生可以自由调阅和交换各种信息,不断提高了医院的整体工作效率,这使得医生对图像进行处理变的更为方便。 医学成像设备的不断发展,检测手段的不断更新,为医生提供了各种不同格式的图像,但是这些图像成像原理不同,反映的信息也很有限,而且大多也只是把三维的人体器官以二维的形式呈现出来,如果能对这些图像进行综合的处理,将使得诊断更准确,因此这也成为医生的迫切需求。 图像后处理技术是利用计算机技术和图像信息理论对成像后的图像来进行各种处理。二维的图像后处理手段如图像增强,通过平滑、锐化处理或者去处噪声或者使边缘清晰;而伪彩处理则可把单调的灰度图像转换成彩色的图像显示出来;图像分割则对不同图像实现对解剖结构或感兴趣区域的提取和描绘。三维的图像后处理则包括各种面绘制和体绘制的三维重建,这可以把一序列的二维医学断层图像重构成三维模型显示出来,提供了更丰富的解剖结构信息,让医生对待检组织和病征有更立体更全面的认识,给出更科学的诊疗方案。配准融合也是一种复杂的后处理手段,可以把多幅医学图像互补的融合成一幅图像,但是要达到较好的应用还需要不断的成熟。 本文主要介绍了PACS平台的特点和应用情况,详细讨论了图像平滑、图像锐化、伪彩处理、图像分割、三维重构、配准融合等各种图像后处理技术的理论和算法,最终设计并实现了一个三维医学影像后处理软件系统。该系统基于PACS平台,将提供各种图像后处理操作,帮助医生优化医学图像,提取更多的有用的信息,提高了诊疗水平。更多还原

【Abstract】 As an information platform for hospital numeralization, PACS (Picture Archiving and Communication System) become more and more important. Currently many hospitals pay much attention to PACS and carry it into execution. According to different hospitals with different scales and needs, PACS can be divided to three types: Full PACS , Department PACS and Mini PACS. PACS always includes information registration, picture viewing, center controlling and other modules. With PACS, doctors can attemper and change information freely, this can improve the hospital’s work efficiency and is very convenient for doctors to deal with pictures.With the development of medical imaging device and measurement technique,there are many kinds of medical images. Because of their different imaging principle, they will display different information. Anyway, most of them are 2D not 3D for the solid apparatus. So if we can deal with these images calling for more information, doctors can make diagnosis more accurately. This becomes doctor’s urgent need.With computer technical and image information theory, Image background processing technical does much treatment to different images. 2D processing includes many means. Image enhancement just like smoothing and sharpening will wipe off yawp or make edge more clear. Pseudocolor processing will change the gray level image to color image. Segmentation will pick-up and display the interested region of different images. 3D processing includes many kinds of surface rendering and volume rendering. These methods can rebuild a series of 2D medical images to a 3D model which will offer much useful information of anatomical structure. Doctors can get a whole acquaintance to the organ and give a proper cure scheme. Image registration and fusion is a complicated processing technical, It can fuse several pictures into one picture, but it is immature for application for the moment.This paper mainly introduces characteristic and application status of PACS, and then gives a discussion about the image smoothing, image sharpening , pseudocolor processing, image segmentation, 3D reconstruction, image registration and image fusion. It gives out these technical’s theory and algorithm. At last we design and implement a 3D medical image background processing software system. This system is based on PACS;itwill provide all kinds of background processing methods. It can help doctors optimize medical images and pick up more useful information. It will improve the diagonosis and treat level.更多还原