Latest News – “Into the Pathoverse” webinar available on-demand

Watch the Video
×
Log in to PathPresenter Public
Log in to PathPresenter Public

Why a Vendor-Agnostic IMS Is Key to a Future-Proof AI Strategy

by Patrick Myles, MBA
CEO, PathPresenter

As AI adoption accelerates in digital pathology, many institutions are facing a strategic decision that will shape their technology stack for years to come: how tightly should AI be coupled to the core imaging and workflow platform?

At PathPresenter, we’ve taken a very deliberate position. On the AI side, we are intentionally AI-agnostic at the IMS and workflow level. That decision is rooted not in theory, but in hard-earned lessons from the field.

The Hidden Cost of Closed AI Ecosystems

Over the years, we’ve seen multiple institutions make large investments in closed scanning and workflow ecosystems, often driven by early regulatory approvals or the promise of an “end-to-end” solution. In the short term, those systems worked. But over time, they frequently became bottlenecks.

When institutions later wanted to onboard additional scanners, integrate third-party tools, or adopt stronger algorithms from competing vendors, those closed platforms made change slow, expensive, or in some cases impossible. The result was delayed deployments, integration failures, and ultimately costly replacement projects that disrupted operations far more than anyone anticipated.

That model simply doesn’t hold up in today’s AI landscape.

AI Is Changing Too Quickly for Locked-In Platforms

AI is moving at a pace we’ve never seen before. Many algorithms deployed in production today are CNN-based, while the next wave of capability is clearly shifting toward foundation-model approaches. As performance improves and new vendors emerge, the “best” solution will continue to change.

In that environment, locking the core IMS and workflow layer to a specific AI vendor is a strategic risk. Our view is simple: the IMS should remain independent of the AI layer.

A Future-Proof Strategy that Keeps the Platform Stable While AI Evolves

To make that separation practical, we’ve invested heavily in building an AI middleware layer designed to be future-proof. This middleware can ingest, normalize, and orchestrate the clinical and operational data required to deploy AI at scale, including integrations built on HL7, FHIR, and DICOMweb. By abstracting these complexities away from the core platform, institutions can adopt new AI models and workflows over time without needing to rebuild integrations or re-platform each time the ecosystem evolves.

Just as importantly, this approach is informed by real-world experience. We’ve worked with a wide range of institutions and organizations and have seen multiple generations of digital pathology and AI vendors come and go. Some early companies were highly innovative, but struggled to scale or keep pace with market shifts.

Those experiences have consistently reinforced the same conclusion: workflow platforms must remain stable, while the AI layer must remain flexible.

In an AI landscape defined by constant change, future-proofing isn’t about predicting which algorithm will win. It’s about designing systems that can evolve, without disruption, as the technology does.

About the Author

Patrick Myles is CEO of PathPresenter. Previously, he was CEO of Huron Digital Pathology, and Vice President of Business Development for Teledyne DALSA. He served as a board member of the Digital Pathology Association.