PACS components such as image viewers and archive offerings are available independently in the market and offer capabilities that rival those of commercial turnkey systems, said Anne Krok, who spoke during a scientific session at RSNA 2013 in Chicago.
"Areas exist in which [these components] may be suitable to bring these functionalities to the desktop in a shorter time compared to commercial vendor turnkey systems," she said.
The benefits of commoditization
In essence, imaging facilities are benefiting from the commoditization of PACS technology. Commoditization is the ability to acquire a desired product by purchasing low-cost components rather than a complete product, Krok said. For example, it's cheaper to buy ingredients and cook a meal yourself than to eat at a fine French restaurant, and it's cheaper to buy tires at a tire dealer than to buy them as part of a complete car.
Support costs are generally far higher for a turnkey system than for the sum of the individual components and in-house trained IT staff, she said. Care and attention need to be given, however, to ensure that all necessary elements are incorporated in the implementation plan.
Robustness of the final system is likely to be higher in this type of model, with markedly lower support costs.
Krok and colleagues sought to assess the advantages and disadvantages of at least partially implementing this approach, focusing on PACS (radiology and nonradiology) and RIS. They evaluated seven components: viewer, archive, worklist or workflow manager, 3D/volume rendering, computer-aided detection (CAD) software, decision-support tools, and engines or tools for integration/interoperability/data exchange.
They studied software offerings and system components from commercially available products. Krok and colleagues also included additional functionalities deemed useful in practice but not available as part of an existing commercial turnkey PACS.
From their analysis, the researchers concluded that the majority of commercial turnkey systems have viewers that lack essential tools. However, separate viewers do exist that offer DICOM connectivity to existing archives and may be implemented to suit specific needs and data types in real-time.
"Different viewers can be implemented simultaneously for different datasets or user needs," Krok said.
Vendor-neutral DICOM archives are the system component that is being marketed most aggressively today, and many options exist in this space.
"Pay attention that the best DICOM archive might well be [from] a vendor that does not supply other commodity components such as a viewer," she said.
A standalone worklist manager is also available as a commercially available product.
"Imagine being able, in real-time and onsite, to create worklists and sorts based upon a specific parameter that exists in a database, such as pregnancy status or even a unique allergy," she said.
Information system needs are no longer rigidly defined by clinical department, Krok said. For example, radiology routinely requires data from other clinical departments, such as the electronic medical record for referrals, lab data/pathology reports, and cardiology.
The workflow of ordering and performing imaging exams may require information normally housed in other traditional information systems; integration engines may be needed to support this model, she said.
Is it practical?
Separate tools do exist to provide 3D/volume rendering capabilities, but desktop integration into viewers is variable, according to Krok.
She said that no consensus has been achieved within radiology regarding the effectiveness and necessity of CAD tools. In addition, the availability and ease of incorporating these tools into a viewer is also variable.
"Server-based CAD tools offer DICOM integration with PACS using DICOM [Structured Reporting]," she said.
Decision support is another area in which many of the necessary tools are likely still lacking for a component-based approach.
"Open architecture can only help in this issue, but this is unavailable from virtually all turnkey vendors," she said.
While it's theoretically possible to acquire every function needed for an information system from a different vendor, the integration and maintenance required makes this approach unpractical.
"The commoditization dilemma renews the considerations between best-of-breed versus best-of-suite," Krok said.
However, specific goals or needs may be best addressed by using a standalone viewer that can accommodate a specific data or image type and is being run by a separate workflow manager working off a vendor-neutral archive.
"If we wait for our large vendors to supply all of our needs, this will take a very long time," she said.
Decisions regarding what to purchase by component and what to acquire via a turnkey system should be a main focus for institutions.
"Carefully consider the support options and cost, including in-house or consultant IT support," she said. "Commoditization is already here."
Copyright © 2014 AuntMinnie.com