[A solution to common failures of the PACS]

Area of Science:

• Medical Imaging Technology
• Computer Engineering
• System Reliability

Context:

• Clinical Picture Archiving and Communication Systems (PACS) often suffer from frequent failures.
• Single points of failure (SPOFs) in existing PACS architectures limit their operational availability.
• High availability is critical for clinical workflows and patient data management.

Purpose:

• To design and implement a Continuous Availability (CA) image server for PACS.
• To enhance system reliability by employing Triple Modular Redundancy (TMR) at the CPU/memory level.
• To mitigate failures and increase the overall availability rate of PACS.

Summary:

• A CA image server utilizing TMR is developed for failover at the CPU/memory level.
• A TMR voter detects failures, ensuring seamless operation.
• Redundant SCSI interfaces connect mirror-image disks, RAID, and DLT controllers for a complete CA system.

Impact:

• The TMR CA image server replaces SPOFs in PACS, achieving a 99.999% availability rate.
• This solution is highly suitable for large-scale medical image networks and database applications.
• Improved system uptime ensures uninterrupted clinical workflows and data access.