From glass to gigapixels: A GI pathology lab’s journey into digital pathology

When I joined OmniPathology, an independent gastrointestinal (GI)–focused pathology laboratory, a little over a year ago as Vice President of Science & Technology, one of the first initiatives I was eager to explore was the adoption of digital pathology. Like many independent laboratories, we had been watching the space carefully for years. Digital pathology had long promised efficiency and scalability, but until recently, the technology, regulatory environment, and cost structure made widespread adoption challenging for a traditional, high-volume anatomic pathology practice.

Over the past few years, however, digital pathology has advanced dramatically. Improvements in whole-slide imaging, faster scanners, more sophisticated image management systems (IMS), and most notably, the rapid maturation of artificial intelligence (AI) tools have fundamentally changed what is possible. AI-enabled digital pathology platforms can now automatically identify regions of interest, perform measurements and cell counts, flag potential diagnostic features, and streamline case navigation in ways that were unimaginable even a decade ago. These capabilities have transformed digital pathology from a promising concept into a practical, powerful, operational tool.

Why digital pathology—and why now

OmniPathology is built on an exceptional foundation: a team of seasoned, world-class, GI fellowship–trained pathologists who collectively have interpreted more than one million biopsies. Over the years, the laboratory developed a highly effective workflow with strong communication between histology, laboratory operations, and pathologists. The rhythm of the lab worked, and it worked well.

At the same time, we recognized that our existing model, like that of many traditional pathology practices, had inherent limitations. Scaling volume required proportional increases in physical space, microscopes, and on-site staffing. Expanding into new subspecialties or service lines could take months or years. Geographic constraints limited how and where pathologists could contribute.

To take advantage of new opportunities, both within GI pathology and in adjacent or emerging diagnostic services, we needed a workflow that was more scalable, more flexible, and less constrained by the physical realities of glass slides and microscopes. Digital pathology offered a path forward.

I was tasked with leading the evaluation process, with the goal of selecting a digital pathology system within a few months. While I had experience with digital tools and laboratory informatics, I had never evaluated or implemented an end-to-end digital pathology ecosystem. My first step was to seek advice from people who had already traveled this road.

Learning from experts

I reached out to two trusted colleagues with deep, real-world expertise: Dr. Eric Glassy, a true pioneer in digital pathology, and Lisa-Jean Clifford, a seasoned leader in high-technology healthcare information solutions. Their guidance was invaluable, helping me understand not only the technology itself, but also the operational, cultural, and strategic implications of going digital.

They emphasized several key principles that shaped our approach:

• Be explicit about your goals and expectations—“digital pathology” means very different things to different laboratories.
• Develop a clear list of critical evaluation criteria and prioritize what matters most to your organization.
• Remember that digital pathology is an ecosystem, not a single product. Scanners, image management systems, viewers, storage, LIS integration, and monitors must all work together.
• Evaluate each component independently and as part of the whole system.
• Bring vendors on site for in-laboratory demonstrations using real cases.
• Involve all stakeholders early—histology, IT, pathologists, and operations staff understand workflow challenges better than anyone.
• Ensure compatibility and ease of integration among the digital pathology platform, image management system (IMS), laboratory information system (LIS), and display hardware.

Armed with this framework, we began a structured, methodical evaluation.

Understanding the digital pathology ecosystem

At its core, any digital pathology solution consists of three primary components:

1. Digital pathology scanners that convert glass slides into high-resolution whole-slide images.
2. An image management system (IMS) that stores, organizes, displays, and analyzes those images while interfacing with the laboratory information system (LIS).
3. Diagnostic-grade monitors that allow pathologists to view digital slides with confidence.

Within this landscape are both FDA-cleared and non–FDA-cleared platforms. FDA-cleared systems provide regulatory reassurance but are often more “locked down,” making customization and rapid innovation more difficult. In contrast, more flexible platforms may allow faster integration of new AI tools and workflow optimizations but require careful validation and governance.

Evaluating digital pathology scanners

Modern scanners have improved significantly in image quality, resolution, and reliability. As a result, throughput and operational resilience became major differentiators in our evaluation. Key criteria included the following:

·       Scanning speed, image quality, and resolution

·       Slide capacity and degree of automation

·       Ease and reproducibility of slide loading

·       Reliability and redundancy in case of downtime

·       Availability and responsiveness of service support

·       Service and maintenance costs

·       Manufacturing location and supply chain considerations

·       Physical footprint in the lab

·       Price and vendor references

Throughput was particularly important. To meet our goal of having cases ready for pathologist review each morning, scanning could not become a bottleneck.

The central role of the image management system

If scanners are the engine of digital pathology, the IMS is the nervous system. It is the software environment in which pathologists review, analyze, and sign out cases, often with AI-assisted tools layered into the workflow.

We paid close attention to:

·       Compatibility with multiple scanners

·       Ease of integration with our existing LIS

·       Robust and high-speed network infrastructure, storage architecture, bandwidth, and scalability

·       Short- and long-term image storage options and pricing

·       Availability and maturity of AI tools

·       Flexibility to integrate third-party or internally developed AI algorithms

·       Customization of the pathologist user experience

·       Vendor location, support model, and long-term roadmap

·       Cost and references

Whole-slide images are large, often around 1 GB per slide, making storage strategy a nontrivial consideration. Most vendors offer tiered storage models, balancing cost against retrieval speed. Selecting the right approach requires a clear understanding of clinical needs, client expectations, and regulatory requirements.

Trial testing in the real world

Rather than relying solely on vendor demonstrations, we arranged week-long, in-lab trials for each scanner under consideration. Instruments were installed sequentially, allowing our histology staff and pathologists to work with each system using real cases and real workflows.

This hands-on experience proved invaluable. Technologists evaluated slide loading, scanning reliability, and ease of use. Pathologists compared digital images directly with glass slides, assessing image fidelity and diagnostic confidence. Reviewing platforms back-to-back helped clarify differences that might otherwise have been subtle.

Similarly, IMS vendors were invited to present their platforms in depth, focusing on workflow integration, customization, AI capabilities, and long-term flexibility. These sessions sparked productive discussions about future use cases that extended well beyond our initial implementation.

Validation: Building diagnostic confidence

Validation is a critical step in implementing digital pathology for primary diagnosis. The College of American Pathologists (CAP) provides clear guidelines for validating whole-slide imaging systems. In essence, selected cases are reviewed on glass and then re-reviewed digitally by the same pathologist after a washout period of at least two weeks. A concordance rate of at least 95% is expected.

Equally important is ensuring that validation covers the full spectrum of cases the laboratory anticipates encountering. For a GI-focused practice, this meant carefully selecting representative biopsies and diagnostic scenarios to ensure digital interpretations were fully commensurate with glass slide review.

Redesigning the workflow

Adopting digital pathology is not simply a technology upgrade, it is a workflow transformation. Our objective was to ensure that all cases could be accessioned, processed, stained, scanned, and ready for digital review by the following morning.

Rather than imposing a top-down solution, we empowered our laboratory team to redesign the workflow from the ground up. They were encouraged to think creatively, without preconceived constraints around shifts, staffing patterns, or traditional processes.

The result was an impressively streamlined workflow that improved efficiency, enhanced scalability, and positioned the lab to absorb future growth and new service lines with minimal friction.

Training and change management

Training is essential to success. While many pathologists adapt quickly to digital review, confidence builds over time. Comprehensive training for histology staff, IT personnel, and pathologists ensured that everyone understood not only how to use the system, but how to troubleshoot and support one another.

We designated several in-house experts to serve as first-line support for the pathology team. This internal safety net made a significant difference, reducing anxiety and accelerating adoption.

Looking ahead

Evaluating and implementing a digital pathology system is a substantial undertaking, particularly for an independent laboratory with established workflows. For OmniPathology, the process proved to be an opportunity to reassess our priorities, modernize our operations, and position ourselves for sustainable growth.

Digital pathology has given us a more efficient, scalable, and forward-looking platform, one that enhances diagnostic quality today while opening the door to AI-driven innovation tomorrow. Access to experienced mentors and resources from organizations such as the College of American Pathologists and the Digital Pathology Association was instrumental in our success.

For laboratories considering the leap into digital pathology, the journey is challenging, but the destination is well worth the effort.