Guest column: The impending deconstruction of PACS
Jim Philbin - 1.05 Mb Network and computing technology has evolved to the point where desktop computers and wired connectivity are becoming relics of the past and the focus has shifted to interoperability and anytime/anywhere information availability (Google, Amazon, Facebook, Netflix).

Customization at the individual user level has become a dominant theme in our socially connected cyber culture. Similarly, cloud technology, deployed around the world, is transforming our culture and is poised to create dramatic changes in healthcare, turning data into meaningful intelligence that saves money but, more importantly, saves lives.

Today, healthcare institutions are challenged to move critical patient data across enterprises. At the same time, they find themselves locked into proprietary vendor agreements, perpetual software licenses that have little to no value after five years and the burgeoning challenge of maintaining or migrating the “non-standard” standard DICOM of their vendors. Meaningful use has put interoperability, innovation and affordability back in the spotlight. They are achievable through the deconstruction of traditional PACS.

Deconstructing PACS?
The medical imaging domain is shifting to a new paradigm that deconstructs the components of PACS into separate applications for visualization, workflow and archiving. By returning to the DICOM standard, especially the new web-based parts, this deconstruction will allow more specialized applications that improve visualization and workflow while making image access and sharing more efficient within and across healthcare systems. Deconstructing PACS will enable faster innovation, elevate vendor accountability and provide imaging solutions for underserved areas such as pathology and dermatology.

PACS was invented more than 20 years ago to manage medical images. Initially, PACS was designed to handle a single type of workflow and the visualization needed for that workflow in a monolithic fashion. It was a great solution at the dawn of digital imaging, but over time, it has become an impediment to advancing access to medical imaging information.

Today, PACS performs three fundamental tasks: stores imaging studies; manages workflow; and controls access to and display of medical images. These functions are interdependent and are traditionally engineered together in one product.  

Today, most systems utilize a client-server model with clients displaying images and worklists while the servers orchestrate workflow, as well as manage and store the imaging studies. PACS communicate with modalities using DICOM, but typically communicate with their clients and storage devices using proprietary protocols.

The recent proliferation of specialized PACS in nuclear medicine, ultrasound, mammography and cardiology—all requiring different workflows and viewers—is further evidence that the single vendor approach is ineffective. The ensuing challenges of integration and interoperability among these systems demonstrate the current limitations of the monolithic PACS.

Moving toward deconstruction
  • The deconstruction of legacy PACS is underway and gaining traction:
  • The DICOM standard is evolving to become more efficient, specifically to facilitate web access and delivery of medical imaging studies;
  • Cross-enterprise image sharing services are addressing specific clinical and financial issues associated with siloed medical image information;
  • Enterprise PACS Vendor Neutral Archive (PVNA) applications are becoming a critical element that eliminate departmental storage and facilitate image access via EMR and portals;
  • Clinical viewing applications, particularly zero footprint clinical viewers, are utilizing server-rendering technology to provide universal access via web browsers or mobile technology and improve IT operations; and
  • The introduction of cloud-based application delivery and support lends a practical and scalable method to reduce costs associated with storage and computing power.

The next steps in this evolution are diagnostic viewers that are independent but connected to PACS or PVNA, along with the delivery of advanced clinical 3D solutions and voice recognition technology as discrete elements of the diagnostic workflow, each interoperable and designed to empower the physician at the point of care.

The last step in the march to deconstructing PACS is the workflow engines themselves, which must be decoupled from viewing and archiving, thus enabling sophisticated workflows within and across institutions. Today, most workflow engines are embedded inside the PACS or RIS. Deconstructing PACS means developing workflow engines that can be easily customized for disparate workflows in different areas of medicine. They will be web-based and enable interchangeable integration with viewers and archives.

Separating the three main components of PACS (viewers, workflow and archive) will create an ecosystem that will allow faster innovation in all three. The increasing use of web-based technologies will reduce the cost of innovation and improve interoperability—the original intent of the DICOM standard. The deconstruction of PACS will lead to increased provider efficiency, better aggregation of medical records and improved patient care delivery.

Jim Philbin, PhD, is former senior director of medical imaging informatics at Johns Hopkins Medicine in Baltimore.