For most of its history, PACS was one product from one vendor: the archive, the database, the viewer and the workflow engine all came in the same box. That model is giving way to enterprise imaging, where images from radiology, cardiology, pathology, dermatology and the point of care are managed as one organisational asset instead of departmental silos.

The vendor-neutral archive

A vendor-neutral archive (VNA) stores images in standard DICOM, and non-DICOM content with standard metadata, independent of the viewing application. Its value shows up on the day you replace a PACS: if the archive is yours and standards-based, a viewer change is a project, not a full data migration. A VNA also gives every department one place to apply retention rules, lifecycle management and patient identity reconciliation.

DICOMweb changes integration

Classic DICOM networking (C-STORE, C-FIND, C-MOVE) is reliable but awkward to build web applications on. DICOMweb exposes the same model over HTTP:

Because it speaks HTTP and JSON, DICOMweb fits modern web viewers, AI pipelines and cloud services without a gateway translating every request. Most current archives and viewers support it; it is worth making a hard requirement in any new procurement.

Zero-footprint viewing

Browser-based diagnostic and clinical viewers remove desktop installs and make images reachable from the EMR, a referring physician's office or home reading. The trade-offs are real: diagnostic reading still needs calibrated displays, fast networks and careful attention to how large datasets stream. For clinical review across the enterprise, though, a zero-footprint viewer launched in context from the EMR is now the expectation.

Cloud, with eyes open

Cloud archives and cloud-hosted PACS are mature options. Their strengths are elastic storage, off-site disaster recovery and less hardware to run. Before committing, model the costs that are easy to miss:

Deconstructed PACS

A deconstructed PACS assembles the archive, viewer, worklist and workflow engine from separate best-of-breed components. It offers flexibility and avoids lock-in, but it moves integration and support onto your team. It suits organisations with a strong imaging informatics group and a clear architecture; for others, a well-integrated single platform on top of a standards-based archive is the safer choice.

What does not change

Whatever the architecture, the fundamentals decide success: accurate patient and study identity, consistent DICOM tagging, tested downtime procedures, and a migration plan that is validated study by study instead of assumed. The technology has moved on; the discipline still matters most.