区域影像信息系统的资源调度研究

摘要

目的:资源调度是区域影像信息系统的核心问题.比传统的中心式区域影像信息系统相比,基于数据网格的区域影像信息系统具有更大的优势.本文拟提出一种区域影像信息系统的资源调度算法的设计与实现,以避免中心资源形成系统瓶颈.方法:首先,本文简单介绍了基于数据网格的区域影像信息系统构架的特点,通过与中心化区域影像信息系统构架的比较,证明其在区域医疗建设中的优势及巨大发展潜力.然后,针对网格区域影像信息系统中核心的科学问题,即资源调度问题进行了分析并提出了基于临床路径过程的均衡调度策略并实现.网格模拟器GridSim被用来测试本文所提出的均衡调度策略的有效性.结果:通过与一种现有的资源-时间均衡的调度算法进行比对,实验结果表明本文所提出的算法能更有效地分配区域影像网格的中心资源,并能提高影像网格对临床紧急事件的响应速度.结论:本文所提出的资源调度算法在调度优先级计算中引入了临床路径知识,能较好地区域影像信息系统中资源的分配问题.%Resource scheduling is one of the key issues in regional medical imaging systems. In comparison with data-center structured imaging systems, grid-based imaging systems have advantages in resource scheduling. In this paper, we propose and implement a resource scheduling method of Grid Picture Archiving and Communication System (PACS) server. Methods: Firstly, a breif introduction of Grid PACS architecture is made. In comparison with conventional PACS architecture, advantages of Grid PACS architecture are displayed in building regional medical information system. Central resource's bottleneck is the key issue of Grid PACS architecture. Then we study on it and propose a new resource scheduling method based on clinical pathway execution. Results: Grid simulation software GridSim was adopted for numerical evaluation of the proposed method. Based on the result comparison with a Cost-Time Optimization Algorithm, it was shown that the proposed method could assign the central resource of Grid PACS server effectively and increases the speed of responses to clinically emergency events. Conclusions: By introducing clinical pathway, our proposed method is better to solve resource scheduling problems in current Grid PACS systems.