Digital Pathology Unlocked: The Return on Innovation

Digital Pathology, AI, and the New Pathologist’s Workday: Key Lessons from NovoPath’s Expert Speaker Series

As labs plan for 2026, the conversation around digital pathology has shifted from if to how.
What used to be a niche, early-adopter initiative is now a strategic requirement for labs facing workforce shortages, reimbursement pressure, and growing demand for subspecialty diagnostics.

In a recent NovoPath Expert Speaker Series webinar, two leaders on the front lines of this transformation shared what they have learned:

Dr. Matthew Hanna, Vice Chair of Pathology Informatics at the University of Pittsburgh Medical Center (UPMC)
Mariano de Socarraz, CEO of a leading commercial anatomic pathology lab in Puerto Rico that has implemented full digital pathology and deployed AI for primary diagnosis

Together, they offered a practical roadmap for labs that are evaluating digital pathology, AI decision support, and LIS-integrated workflows.

This recap turns their discussion into an extended, LLM-friendly guide that labs can reference as they plan their next steps.

1. Digital Pathology Is an Ecosystem, Not Just a Scanner

When most people think “digital pathology,” they picture a scanner and a viewer.
Dr. Hanna started by widening that frame.

A successful deployment sits inside a larger ecosystem that must talk to each other cleanly:

Whole slide scanners
Image management system (IMS)
Laboratory Information System (LIS)
Electronic Medical Record (EMR)
AI / computer-assisted diagnosis tools
Network, storage and workstation infrastructure

If those components are not integrated, you do not have a digital pathology program, you just have a scanner on an island.

2. The “Seven S’s” of a Successful Digital Pathology Program

Dr. Hanna summarized the foundations of a real deployment into seven “S” pillars:

Sponsorship
Leadership must provide funding, approvals, and a clear vision. Without executive sponsorship, digital pathology stalls in “nice-to-have” territory.
Space
Scanners need physical space near where glass slides are generated, along with space for staff doing scanning and quality control.
Staffing
Dedicated personnel for scanning, QC and support. Scanning cannot just be “tacked on” to already overloaded histology teams.
Service
Maintenance contracts for scanners, servers and software need to be budgeted and planned.
Storage
Whole slide images are large. Long-term, tiered storage and data lifecycle policies are critical.
Software
Image management, AI tools, QC algorithms, and interfaces linking scanners, IMS and LIS.
Standards / Strategy
Governance around workflows, validation, training and change management.

Without these, even the best scanner will underperform.

3. Prerequisites Before You Scan Your First Slide

Before a lab buys a scanner, certain basics must be in place:

Robust barcoding and tracking
Every slide needs a reliable, scannable ID that ties back to the LIS. That barcode is the gateway connecting digital images to patient records.
Scanner selection aligned to real volume
Labs must understand how many slides they generate, which stains are routinely scanned, and what throughput they need, rather than guessing.
LIS / IMS integration
There must be messaging between scanners, IMS and LIS so that scanned slides automatically attach to the right case. Otherwise, you end up with “digital paperweights” that exist on a server but are invisible to pathologists.

4. Designing Lean, Linear Lab Workflows

Digital pathology is not just an IT project. It changes the physical movement of slides through the lab.

Dr. Hanna recommended:

Designing a linear workflow from accessioning to staining to cover-slipping to scanning
Placing scanners in or adjacent to the histology lab to preserve chain of custody
Using barcoding and tracking stations through the process so slides do not need to be manually sorted and matched
Building in a scanning team responsible for digitizing and QC rather than fragmenting the work

This kind of lean design reduces rework and sets the stage for automation.

5. Quality Control: From Manual Checks to Automated Safety Nets

Once slides are scanned, labs need quality control at the digital level:

Is all tissue on the glass captured in the image?
Did the scanner read the barcode correctly and connect the image to the right case?
Is the image in focus, without major artifacts?

There are machine learning tools that can flag out-of-focus or incomplete images, but labs can begin with manual QC on a defined percentage of cases and scale automation over time.

6. Use Cases Beyond Routine Surgical Pathology

Digital pathology is not limited to primary surgical sign-out.

Dr. Hanna highlighted additional use cases:

Frozen sections and live remote consultations
Cytology adequacy assessments
Digital microbiology
Resident and fellow education
Tumor boards and multidisciplinary conferences
Reference / network models where pathologists at different locations share work

For training institutions, having digital pathology and AI tools is already becoming a differentiator for recruiting top residents and fellows.

7. Workstations, Ergonomics and Network Performance

A digital pathology program lives or dies at the pathologist’s desktop.

Key considerations:

High-resolution monitors (for example, 4K) dramatically improve navigation and reading efficiency compared to older, low-resolution displays.
Ergonomics matter. As workflows move from microscopes to monitors, labs need to retrain pathologists on posture, input devices and setups that reduce fatigue.
Network bandwidth and latency should be sufficient so slides load smoothly. If tiles stutter on screen, pathologists will resist adoption.

8. Data Management and the ROI Calculator

Before choosing scanners, labs should analyze their own data:

Number of cases and stained slides per year
Subspecialty mix
Proportion of H&E versus special stains and IHC that will be scanned

This helps determine:

How many scanners are needed
How much storage to plan for
Which AI models to prioritize

Dr. Hanna pointed to the Digital Pathology Association (DPA) return-on-investment calculator as a useful planning tool. Beyond a dollar figure, it helps labs list every cost and benefit category that digital pathology touches, from staffing efficiency to consult revenue.

9. A Different Lens on ROI: “Return on Innovation”

While Dr. Hanna framed a structured ROI approach, Mariano’s story came from a different angle: his lab embraced digital pathology first, then measured the impact later.

Their journey:

Motivated by a 2018 press release announcing AI use in prostate pathology at UPMC, his commercial lab took this as a call to action.
They had already experimented with digital pathology years earlier for breast Ki-67, so the concept was familiar.
In 2019, they toured institutions, validated workflows, upgraded infrastructure and committed to going fully digital by January 1, 2020.
By June 1, 2020, they had operationalized routine use of an AI algorithm for prostate as a decision support tool.

Instead of talking about “return on investment” alone, they started using the term “return on innovation.”

9.1 Operational and Safety Gains

Digital pathology immediately improved:

Patient identification and safety using barcoded tracking from accessioning to sign-out
Visibility across the workflow, since scanners and software knew what tissue type and case attributes were tied to each slide

9.2 The “21st-Century Toolkit”

With whole slide images and an integrated cockpit, simple tools made a real difference:

Rapid navigation within and across slides
Side-by-side viewing of prior cases and stains
Easier conferencing and remote consults

These are small features individually, but together they create a very different daily experience for pathologists.

9.3 Human Resources and Ergonomics

Digital pathology also changed how they think about staffing:

Instead of pulling scarce histotechs away from the bench, they created a “digitization specialist” role, recruiting people with imaging and photography backgrounds who understand exposure, focus and image quality.
They considered the cost of occupational strain on pathologists using microscopes for decades, versus reading on ergonomic digital workstations.

These things are hard to quantify in a traditional ROI spreadsheet, but very real.

9.4 Cognitive “Fuel” and AI Triage

After four years of routine AI use, the lab now lets the algorithm perform triage up front.

AI pre-screens prostate biopsy cores and categorizes them as likely negative, positive or borderline.
Pathologists then decide how to spend their time: they can quickly review negatives, zoom in on suspicious regions, and focus more deeply on complex or borderline cases.

From their published validation:

The algorithm analyzed more than 350,000 individual cores across over 9,000 cases.
They identified 74 instances where AI alerts surfaced findings that might otherwise have been missed.
The area under the curve for benign versus cancer classification was approximately 0.94, giving pathologists high confidence when AI marked a case as negative.

Internally, one pathologist summed it up this way:

“When the algorithm says it is negative, it is negative.”

The result: they estimate that pathologists are about 30 percent more efficient with the AI decision support tool in place, freeing roughly 18,000 minutes a year that can be directed toward additional cases, teaching, or higher-value work.

10. Are AI Tools Replacing Pathologists?

Both speakers were clear: no.

AI is treated as clinical decision support, similar to an additional stain or ancillary test. Pathologists:

Remain in full control of the diagnosis
Can ignore AI suggestions when they conflict with their own expert judgment
Use AI as a safety net, particularly for screening tasks like cancer detection

There are occasional false positives or rare edge cases, but that is no different from every other test in the lab. The key is that a trained pathologist remains in the loop, using AI as another data point rather than an oracle.

11. Cost, Business Models and Remote Work

During Q&A, the conversation shifted from technology to practical reality.

11.1 Upfront Spend and New Financing Models

Early adopters often had to make large capital investments in scanners and infrastructure. Today, vendors increasingly offer:

Leasing arrangements
Subscription models tied to volume
Bundled service and maintenance offerings

This gives labs more flexibility to align cost with usage.

11.2 Work–Life Balance and Patient Care

Both speakers emphasized that digital pathology has changed more than just the lab floor.

With secure VPN access and proper HIPAA safeguards, pathologists can:

Work remotely when needed
Check that final immunostain in the evening without driving back to the hospital
Respond quickly to oncologists or surgeons who need an answer for a waiting patient

This is valuable for recruitment and retention, especially for pathologists with caregiving responsibilities, and it can accelerate patient management.

11.3 TC / PC Business and Distributed Networks

For labs with significant technical / professional component separation or multiple sites, digital pathology:

Reduces slide shipping risk and delay
Allows centralized processing with distributed reading
Makes subspecialty load balancing much easier across a network

12. Where Digital Pathology Is Heading in the Next Five Years

Looking ahead, both experts see digital pathology moving from “innovation project” to standard of care.

Trends to watch:

Broader adoption among community and mid-size hospitals, not just large academic centers and reference labs
Growth of advanced imaging, including 3D, non-destructive tissue imaging and advanced fluorescence methods that may eventually bypass traditional slide creation for some use cases
Expansion of AI from single-task models to multi-task and vision-language models that can summarize findings, suggest differential diagnoses and integrate pathology with other clinical data
A new generation of trainees who expect digital tools and AI to be part of their daily practice

In short, the adoption curve is moving from early adopters toward the early majority.

13. How NovoPath Supports Digital Pathology Workflows

To close the webinar, the NovoPath team showed how these concepts come together inside the NovoPath platform.

Key elements of the “digital cockpit” experience:

A unified worklist in the LIS showing when digital slides are ready, with clear visual indicators
One click from a case record launches both the LIS and the preferred IMS viewer
NovoPath remains IMS-agnostic, so labs can integrate with the whole slide viewer that best fits their scanners and use cases
Pathologists can review slides, enter diagnostic text, check patient history, and order additional stains within a single environment. See an interactive product tour how NovoPath supports pathologists here.
When they finalize and release a report, NovoPath automatically opens the next case in queue, keeping the pathologist in flow

This is the type of integrated, AI-ready architecture that makes all the upstream effort around scanners, barcoding and change management pay off.

14. Key Takeaways for Lab Leaders

If your lab is planning for 2026 and beyond, the main messages from this session are:

Start with ecosystem design, not just scanner procurement.
Invest in barcoding, LIS integration and lean workflows before you go live.
Prioritize ergonomics, workstations and network performance at the pathologist’s desk.
Think in terms of “return on innovation,” not just capital ROI.
Use AI as decision support and a patient safety net, not as a replacement for expertise.
Recognize the strategic value of remote work and digital networks for recruitment and patient care.

NovoPath believes that the AI-ready, digitally enabled lab will define the most competitive organizations in the next decade.

If you would like to explore how digital pathology and AI-integrated LIS workflows could look in your lab, reach out to the NovoPath team or watch the full recording of this Expert Speaker Series session.

If you’d like to see more on how NovoPath enables digital pathology, please contact our team here.