Performance evaluation of a picture archiving and communication network using stochastic activity networks

Area of Science:

• Medical Imaging Technology
• Computer Networking
• Telecommunications

Background:

• Picture Archiving and Communication Systems (PACS) are crucial for managing medical images.
• Efficient network design is essential for timely image retrieval and system performance.
• Existing PACS networks face challenges in handling large data volumes and maintaining control flow.

Purpose of the Study:

• To discuss a novel fiber-optic star-based PACS network utilizing wavelength-division multiplexing (WDM).
• To evaluate the performance of this PACS network under realistic workload conditions.
• To demonstrate the suitability of stochastic activity networks for modeling PACS performance.

Main Methods:

• Development of a PACS network with separate image (INET) and control (CNET) channels using WDM.
• Modeling of the INET and CNET using stochastic activity networks.
• Performance evaluation of key variables like image response time and channel delay under various workloads.

Main Results:

• Stochastic activity networks are validated as an effective modeling tool for PACS.
• Specific workload conditions for optimal PACS operation are identified.
• The control channel (CNET) demonstrates a consistently light load, even under high demand.

Conclusions:

• The proposed WDM-based fiber-optic PACS network is a viable solution for medical imaging.
• The study provides insights into the operational limits and performance characteristics of such systems.
• The control network is robust and efficient, ensuring smooth data flow management.