DICOM is both a format and a protocol. Their adoption gave birth to the future of digital imaging, and with it the possibility of remote interpretations in shorter time. As a format, DICOM is a powerful specification with enormous benefits in medicine allowing parties to share cases across multi-vendor environments. As a protocol, DICOM was designed as a robust way to move studies from modalities to PACS to workstations in a hospital or a closed campus, where the network elements (e.g., switches, routers, cables, fiber) are all within the purview of an IT department, When it comes to wide-area networks (WANs), however, i.e., the great expanse between two enterprise networks, DICOM presents a number of significant shortcomings.

The first: DICOM is a synchronous protocol. In this area, the protocol is analogous to faxing where both fax machines must be present for the entire send/receive transaction. Just as one cannot send a fax to a machine that doesn’t answer, one cannot use DICOM to send a study to a device that does not respond. There are many negative consequences of this synchronicity. One of the implications is that both the sending and receiving ends must be “on,” ready, and functioning properly – hardly always a certainty. A second key implication of this synchronicity in DICOM is that the protocol is very “chatty,” requiring significant acknowledgements, and counter-acknowledgments, from the devices sending and receiving using the protocol. This need for constant handshaking creates further delays and can itself be slowed down when factoring the speed at which the sending and receiving devices produce the acknowledgements. This is one of the reasons the concurrent send or receipt of studies is so limited in the DICOM world.

The second shortcoming is statelessness. DICOM has no memory of its state during an active transfer. If, during the transfer of a 1,000 image study, something should happen that interrupts the process even momentarily, DICOM cannot “remember” which was the last image properly received, and starts again from the first one. This has a significant impact on the effective transfer rate of cases between DICOM devices.

While in the local area network (generally 10/100 Mbps or Gigabit Ethernet) the delays caused by the two shortcomings above can be easily tolerated, over the WAN they become exceedingly problematic, making the use of the DICOM protocol over WANs extremely slow. This sluggishness is called DICOM Latency and cannot be greatly reduced by simply compressing and adding bandwidth at the edges of transfer links, i.e., increasing the traffic throughput of the sending and receiving facilities to the Internet.

AG Mednet has developed patent pending technology, aimed specifically at the problem of latency in the transfer of DICOM studies across WANs. DICOM traffic moving over AG Mednet is asynchronous, stateful and can be routed to one or more destinations simultaneously. Additionally, receiving facilities are able to handle more volume through AG Mednet’s high concurrency capabilities.

The combination of AG Mednet’s proprietary traffic grooming methodologies, integrated with its unique routing algorithms all traveling across its expansive network, represent the fastest, most effective platform to move DICOM studies of any size between hospitals, imaging centers and core labs anywhere in the world.