Suppr超能文献

使用带有分层元数据服务器的分布式文件系统的分布式PACS。

Distributed PACS using distributed file system with hierarchical meta data servers.

作者信息

Hiroyasu Tomoyuki, Minamitani Yoshiyuki, Miki Mitsunori, Yokouchi Hisatake, Yoshimi Masato

机构信息

Department of Life and Medical Sciences, Doshisha University, Japan.

出版信息

Annu Int Conf IEEE Eng Med Biol Soc. 2012;2012:5891-4. doi: 10.1109/EMBC.2012.6347334.

Abstract

In this research, we propose a new distributed PACS (Picture Archiving and Communication Systems) which is available to integrate several PACSs that exist in each medical institution. The conventional PACS controls DICOM file into one data-base. On the other hand, in the proposed system, DICOM file is separated into meta data and image data and those are stored individually. Using this mechanism, since file is not always accessed the entire data, some operations such as finding files, changing titles, and so on can be performed in high-speed. At the same time, as distributed file system is utilized, accessing image files can also achieve high-speed access and high fault tolerant. The introduced system has a more significant point. That is the simplicity to integrate several PACSs. In the proposed system, only the meta data servers are integrated and integrated system can be constructed. This system also has the scalability of file access with along to the number of file numbers and file sizes. On the other hand, because meta-data server is integrated, the meta data server is the weakness of this system. To solve this defect, hieratical meta data servers are introduced. Because of this mechanism, not only fault--tolerant ability is increased but scalability of file access is also increased. To discuss the proposed system, the prototype system using Gfarm was implemented. For evaluating the implemented system, file search operating time of Gfarm and NFS were compared.

摘要

在本研究中,我们提出了一种新的分布式PACS(图像存档与通信系统),它可用于整合各医疗机构中现有的多个PACS。传统的PACS将DICOM文件控制在一个数据库中。另一方面,在所提出的系统中,DICOM文件被分离为元数据和图像数据,并分别存储。利用这种机制,由于并非总是访问文件的全部数据,因此诸如查找文件、更改标题等一些操作可以高速执行。同时,由于利用了分布式文件系统,访问图像文件也可以实现高速访问和高容错性。所引入的系统还有一个更重要的点。即整合多个PACS的简易性。在所提出的系统中,仅整合元数据服务器即可构建整合系统。该系统还具有随文件数量和文件大小的增加而扩展文件访问的能力。另一方面,由于整合了元数据服务器,元数据服务器是该系统的薄弱环节。为了解决这一缺陷,引入了分层元数据服务器。由于这种机制,不仅容错能力增强,而且文件访问的可扩展性也增强。为了讨论所提出的系统,实现了使用Gfarm的原型系统。为了评估所实现的系统,比较了Gfarm和NFS的文件搜索操作时间。

文献检索

告别复杂PubMed语法,用中文像聊天一样搜索,搜遍4000万医学文献。AI智能推荐,让科研检索更轻松。

立即免费搜索

文件翻译

保留排版,准确专业,支持PDF/Word/PPT等文件格式,支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述,25分钟生成高质量综述,智能提取关键信息,辅助科研写作。

立即免费体验