Analyzing the load capacities of PACS networks
“When we have a computer system we have many variables,” said Sergio Camorlinga, PhD, of St. Boniface Hospital Research Centre, at a scientific session hosted by the Society for Imaging Informatics in Medicine (SIIM, formerly SCAR) show in Austin, Texas. “This project is focusing on networks and moving data across. We wanted a fast approach to diagnosis, the network and transmission of images.”
The scientific session titled, “A Tool for Analyzing the Load Capacities of PACS Networks,” was presented during a SIIM session focused on infrastructure and image acquisition.
PACS administrators must have a thorough knowledge of the traffic capabilities of their networks to permit the timely, reliable transmission of electronic medical images for radiologists and clinicians to care for their patients. PACS permits the transmission of medical images from modalities to archives, from archives to reading stations and other locations throughout the healthcare enterprise where data are needed.
However, tools can be used by PACS administrators to manage networks efficiently, said Camorlinga, and he introduced a research project at his facility in which non-invasive tools were implemented and utilized for collecting data on a PACS network.
“We used a distributed network of modality simulators called slaves,” said Camorlinga. “The simulators push studies to the PACS and collect response time information. The simulators then communicate this information with a service portion of the program, called master. The master can then compile the information in graphs. These graphs can be used by PACS administrators to better understand how their network responds to different levels of traffic.”
Camorlinga explained that when the statistics are graphed, the plot yields information as to what is the best performance for PACS under a set of given loads. “Test configurations can then be changed to simulate the effects of changes to the network,” said Camorlinga. “An example of this would be simulation and addition of new scanners to the hospital.”
Camorlinga suggested that other healthcare institutions could use the statistical information to see if the transfer times are within the tolerance parameters for the hospital. “With the information, the PACS administrator can quickly see if further upgrades need to be made to the network before additional modalities or reading stations or other applications are added to the network,” he said.
Keeping a PACS running reliably requires proactive measures, recommended Camorlinga. PACS administrators can be prepared ahead of time to handle the changes constantly made to a hospital’s network and information systems, since radiology is a dynamic environment where continuously new modalities, reading stations and other devices are added and or modified.
The scientific session titled, “A Tool for Analyzing the Load Capacities of PACS Networks,” was presented during a SIIM session focused on infrastructure and image acquisition.
PACS administrators must have a thorough knowledge of the traffic capabilities of their networks to permit the timely, reliable transmission of electronic medical images for radiologists and clinicians to care for their patients. PACS permits the transmission of medical images from modalities to archives, from archives to reading stations and other locations throughout the healthcare enterprise where data are needed.
However, tools can be used by PACS administrators to manage networks efficiently, said Camorlinga, and he introduced a research project at his facility in which non-invasive tools were implemented and utilized for collecting data on a PACS network.
“We used a distributed network of modality simulators called slaves,” said Camorlinga. “The simulators push studies to the PACS and collect response time information. The simulators then communicate this information with a service portion of the program, called master. The master can then compile the information in graphs. These graphs can be used by PACS administrators to better understand how their network responds to different levels of traffic.”
Camorlinga explained that when the statistics are graphed, the plot yields information as to what is the best performance for PACS under a set of given loads. “Test configurations can then be changed to simulate the effects of changes to the network,” said Camorlinga. “An example of this would be simulation and addition of new scanners to the hospital.”
Camorlinga suggested that other healthcare institutions could use the statistical information to see if the transfer times are within the tolerance parameters for the hospital. “With the information, the PACS administrator can quickly see if further upgrades need to be made to the network before additional modalities or reading stations or other applications are added to the network,” he said.
Keeping a PACS running reliably requires proactive measures, recommended Camorlinga. PACS administrators can be prepared ahead of time to handle the changes constantly made to a hospital’s network and information systems, since radiology is a dynamic environment where continuously new modalities, reading stations and other devices are added and or modified.