Data Grid for PACS and Medical Imaging Informatics
In Chapter 20, we presented the framework of DICOM-based medical imaging informatics Figure 21.1 shows the topics and concepts covered in Part IV with regard to the five logical layers of the medical imaging informatics infrastructure (see Fig. 21.2). This chapter discusses the Data Grid and its applications in PACS and medical imaging informatics.
21.1 DISTRIBUTED COMPUTING
21.1.1 The Concept of Distributed Computing
Data Grid, which is a component of grid computing, a concept that has evolved from distributed computing. The basic idea behind distributed computing is that if several computers are networked together, the workload can be divided into smaller pieces for each computer to work on. In principle, when n computers are networked together, the total processing time can be reduced down to 1/n of the single-computer processing time. It should be noted that this theoretical limit is unlikely to be achieved because of various unavoidable overheads, for example, data communication latency.
Two important factors affect the design of a distributed computing schema. Processor speed variations in different computers make it important to implement the mechanism to balance the workload in distributed computing in a way that allows faster computers to be given more work to do. Otherwise, the speed of the processing will be slowed down by the slower computers in the distributed computing network. Data communication speed ...