摘 要
随着PACS(图像存储与通信系统)的普及，作为PACS的核心的DICOM(医学数字成像及通信)标准变得越来越重要。DICOM是用于医学图像和通信的国际标准，它规范了医学图像及各种数字信息在异构系统和设备之间存储、传送时的文件格式和语义描述。
论文首先介绍了DICOM 3.0标准，系统地分析了DICOM对象关系和文件格式，主要阐述了数据元和数据集的格式，然后构造了整个阅片工作站的5功能模块。论文提出严格按照顺序读取DICOM字段的做法，尽量减少访问DICOM文件的时间以加快读取DICOM文件信息的算法效率；论文中引入斜率和偏移量这两个概念，简化了VOI LUT（兴趣值查找表）转换的线性插值函数,并采用将实数乘除转化为整数运算的方法简化线性映射函数，从而加快了图像缩放的速度；论文最后详细描述了图像测量和标注用到的类和函数，并分析了具体的测量和标注操作的实现步骤。
论文最终完成了1个具备浏览、缩放、灰度调整、文件保存功能的小型阅片工作站的算法设计，实现了1个小型DICOM阅片工作站。
关键词
DICOM；窗宽和窗位；平滑处理；线性插值
Abstract
With the popularization of PACS (Picture Archiving and Communica- tion System), as the kernel of PACS, DICOM (Digital Imaging and Communication in Medicine) become increasingly important. DICOM is an international standard in medical imaging and communications. It defines the rules of pixel data and other digital information when they are transferred or stored between devices with different operating systems or of different manufacturers.
This paper first introduces the DICOM 3.0 file format, analyzes object connection and file format of DICOM systematically，expatiates Data Set and Data Element mostly and then conformation five modules of Image Browse Workstation. Paper puts forward the method of read DICOM field in term of sequence, reduces the time of read DICOM file in order to mend the arithmetic efficiency of read DICOM file information. Paper introduces two concept, slope and shift, reduces the linear interpolation function of VOI LUT (Value of Interest Look Up Table), and uses the way of make real number operation into integer operation to simplification the linear map function, thereby makes the image zoom speed up. Paper depicts the class and function of image measure and label particular at last, and analyses actualize step of the operation of measure and label.
Paper achieves the algorithm design of a small Image Browse Workstation with the function of browse, zooms, grey level transformation, file save, and achieves a small DICOM Image Browse Workstation.
Keywords
DICOM；Window Width and Window Center； smoothing processing； linear interpolation
