Workflow-Enabled Distributed Component-Based Information Architecture for Digital Medical Imaging Enterprises

Few information systems today offer a flexible means to define and manage the automated part of radiology processes, which provide clinical imaging services for the entire healthcare organization. Even fewer of them provide a coherent architecture that can easily cope with heterogeneity and inevitable local adaptation of applications and can integrate clinical and administrative information to aid better clinical, operational, and business decisions. We describe an innovative enterprise architecture of image information management systems to fill the needs. Such a system is based on the interplay of production workflow management, distributed object computing, Java and Web techniques, and in-depth domain knowledge in radiology operations. Our design adapts the approach of "4+1" architectural view. In this new architecture, PACS and RIS become one while the user interaction can be automated by customized workflow process. Clinical service applications are implemented as active components. They can be reasonably substituted by applications of local adaptations and can be multiplied for fault tolerance and load balancing. Furthermore, the workflow-enabled digital radiology system would provide powerful query and statistical functions for managing resources and improving productivity. This paper will potentially lead to a new direction of image information management. We illustrate the innovative design with examples taken from an implemented system.