PPT3

• PACS infrastructure design concept

  4 major ingredients
• 4 major ingredients in the PACS infrastructure design concept

  • INDUSTRY STANDARDS
  • CONNECTIVITY AND OPEN ARCHITECTURE
  • RELIABILITY
  • SECURITY
• Industry standards to be used in PACS infrastructure design


  • UNIX operating system
  • WINDOWS NT/XP operating system
  • C and C++ programming languages
  • Java programming language platform
  • XML (extensible markup language) for data representation and exchange
  • SQL (structured query language) as the database query language
  • X WINDOW platform for graphical user interface (GUI)
  • TCP/IP communication protocols
  • DICOM standard for image data format and communication
  • HL7 (Health Level 7) for healthcare database information and textual data format exchange
  • IHE (Integrating the Healthcare Enterprise) for workflow profiles
  • ASCII text representation for message passing
• Connectivity and open architecture


  Ensure the PACS allows for future connectivity and transmission of images and data between modules
• Reliability


  • Use fault tolerant measures, including error detection and logging software, external auditing programs, hardware redundancy, and intelligent software recovery blocks
  • Fail recovery mechanisms like automatic retry of failed jobs with alternative resources and algorithms and intelligent bootstrap routines that allow a PACS computer to automatically continue operations after a power outage or system failure
• Security


  Ensure patient confidentiality due to medicolegal issues and HIPAA mandate<|>Application programs may supply additional layers of protection like privilege control to grant and revoke user access to specific database tables, columns, or views
• System architecture


  The hardware and software infrastructure of a computer system
• Stand-alone model


  Also known as "Thick client model"<|>Key features: Images automatically sent to diagnostic and review workstations, workstations can query/retrieve images from archive server, workstations have short-term cache storage
• Stand-alone model data flow

  1. RIS notifies modality and PACS server of patient registration

  2. Modality sends images to PACS server

  3. PACS server archives images

  4. PACS server distributes copies of images to selected diagnostic and review workstations

  5. PACS server prefetches historical images and sends to workstations

  6. Workstations have local storage to hold PACS exams

  7. Workstation returns diagnosis to RIS via PACS server
• Advantages of stand-alone model

  • If PACS server goes down, modalities/acquisition gateways can send images directly to workstations
  • Multiple copies of PACS exams distributed, less risk of data loss
  • Historical exams available on workstation local storage
  • Less susceptible to network performance issues
  • Ability to modify DICOM header during quality control before archiving
• Disadvantages of stand-alone model

  • Rely on correct image distribution and prefetching based on preset defaults
  • Inconvenient worklist management across multiple workstations
  • Need to use complex query/retrieve function to retrieve exams
  • Potential for duplicate readings across workstations
• Client-server model


  Also known as "Thin client model"<|>Key features: Images centrally archived on PACS server, workstations have single worklist to retrieve exams, workstations have no local storage and flush images after reading
• Client-server model data flow

  1. RIS notifies modality and PACS server of patient registration

  2. Modality sends images to PACS server

  3. PACS server archives images

  4. Workstations access complete worklist on PACS server to retrieve images

  5. Images loaded directly into workstation memory for viewing, flushed after reading

  6. Workstation returns diagnosis to RIS via PACS server
• Advantages of client-server model

  • Any exam available on any workstation at any time
  • No prefetching or study distribution needed
  • No complex query/retrieve function needed
  • Avoid duplicate readings across workstations
• Disadvantages of client-server model

  • PACS server is single point of failure
  • More database transactions, less robust than stand-alone
  • Dependent on network performance, problematic for teleradiology
  • No DICOM header modification until image is archived
• Web-based model


  Similar to client-server model but with web-based client and server software
• Advantages of web-based model

  • Platform independent client workstations
  • Completely portable, can be used on-site or remotely
• Disadvantages of web-based model

  • May be limited in functionality and performance compared to client-server model
• Enterprise PACS and ePR (Electronic Patient Record) system


  Very large scale PACS systems integration with centralized data center
• Enterprise PACS and ePR workflow

  1. Patient images and data from each PACS sent to enterprise data center for long-term archive

  2. Filtered patient data and images sent from each site to ePR system in data center

  3. Data center has primary and secondary data centers to avoid single point of failure

  4. ePR system combines patient records with images from all sites

  5. ePR system backed up at secondary data center

  6. ePR web clients throughout enterprise can access patient records and images from any site