Advances in Internet technologies are making wide-area image distribution increasingly practical and affordable for hospitals and imaging centers of all sizes. The concept is not new; teleradiology has been around for years. But the ability to use
Advances in Internet technologies are making wide-area image distribution increasingly practical and affordable for hospitals and imaging centers of all sizes. The concept is not new; teleradiology has been around for years. But the ability to use off-the-shelf personal computers, Web-based browsers, and standard phone lines -- instead of the high-resolution monitors, specialized software, and high-bandwidth connections common to most teleradiology installations -- is helping to make the distribution of digital diagnostic images outside of the radiology department accepted, and even expected.
The trend toward enterprise image distribution has prompted every major imaging and PACS vendor to add standard browsers and other Web-based capabilities to the latest versions of their products. A new market has also been created for software companies, and several now offer Internet-based medical imaging packages that are challenging the more established players and products.This new breed of medical imaging vendors includes Amicas, eMed, Image Medical, Riptide, Stentor, and RealTimeImage. Their products use standard Internet tools such as Java and XML (extensible markup language) and communication protocols such as TCP/IP that only a few years ago were foreign to most device and systems manufacturers.
JAVA TECHNOLOGY
Amicas was one of the first software developers to recognize the advantages of Java and other Web-based development tools for enterprise image distribution. The company was founded in 1995 to commercialize technology developed by Drs. James Thrall and Adrian Gropper (now chairman and chief technology officer of Amicas) at Massachusetts General Hospital. MGH eventually negotiated an equity position in the company, giving Amicas access to all MGH intellectual property in exchange.
Much of Amicas' early business was in on-call teleradiology. Today the company's 59 enterprise image management customers include MGH, Boston University Medical Center, Columbia Presbyterian in New York, the Loma Linda VA, and Kaiser Permanente in Oakland, CA. The company also claims to manage more than two million radiology studies annually.
"The only thing available to replace film back in 1996 was hardware-based proprietary 'islands of information,'" said Barry Gutwillig, vice president of development for Amicas. "PACS vendors today still consider radiology their core business, but they have now put Web server architecture on top. We took the opposite approach: Start with the enterprise, then move downstream to the radiology department."
The Amicas Web/Intranet Image Server is a Java-based software product that uses standard Web browsers (Netscape Communicator and Internet Explorer) to access and view images at flexible resolution levels. The Amicas software is designed to "sense" a receiving monitor's size and resolution and whether multiple monitors are being used and then adjust the file to the appropriate resolution. The Amicas software can also accommodate HL7, SQL, or HTTP requests to bring reports to the viewer. A Java Bean or Active X control allows the viewer to be launched from inside a RIS, HIS, or other application.
"Unlike traditional PACS vendors that bolt on a Web server to their database and archive, the Amicas architecture is built around and on top of a Web server," Gutwillig said. "So instead of multiple boxes with multiple databases, Amicas uses a single database that controls everything from the DICOM importer to the data manager and Web server."
XML TECHNOLOGY
One of the newest entrants to the enterprise image distribution market, Riptide Technologies believes its XML-based technology offers several advantages over the Java-based products developed by Amicas and others. Riptide was founded in 1997 by Dr. Richard Ferrans and Perry Schwalb, who had developed a prototype cardiac ultrasound image capture and Internet delivery system. Ferrans and Schwalb lifted many of the concepts used in building the protoype to develop ARIIS, Riptide's first commercial product.
ARIIS is a Web-based system that combines RIS, PACS, report generation, and charge-capture capabilities in an open-architecture design. The user interface and clinical workstations are based on Riptide's Active DICOM system, which allows images to be displayed and manipulated inside Internet Explorer. In addition, unlike other Web-based systems, ARIIS displays DICOM standard images rather than those compressed with proprietary wavelets.
Riptide was one of the first companies to recognize the implications of XML as a development environment for medical imaging and enterprise distribution applications. For example, the XML framework allows ARIIS to cache images on the user's hard drive or the local network so that existing studies can be automatically prefetched and stored in the background, even before the patient comes in, something that is not possible with Java, Ferrans said.
In addition to being written in XML, ARIIS uses standard Web protocols such as TCP/IP and HTTP to transmit information more efficiently over the Internet.
"A lot of vendors continue to use the DICOM communications protocol for communication between different workstations, which is a very cumbersome method once you get outside of the LAN," Ferrans said. "With TCP/IP and HTTP, you can make the whole world look like the local network."
ARIIS has been commercially available only since Aug. 1, and Riptide can claim just two customers, both of them beta sites. The company is anticipating a boost in exposure from a new agreement with Intel Internet Authentication Services that will allow Riptide to offer advanced Internet security to hospital radiology departments and imaging centers. Riptide will integrate Intel's physician identity authentication services, which utilize Internet IDs issued to physicians by the American Medical Association, into ARIIS.
For more on Web image distribution:
Web Radiology: A new way to store, read, and distribute images (DI, November 1999)