What is Digital Pathology?

• Digital pathology is the acquisition, management and analysis of digitized glass slides in a digital environment.
• Whole slide imaging (WSI) scanners capture an entire slide at high magnification, creating files that can reach tens of gigabytes before compression.
• In 2017, the FDA authorized the first whole slide imaging system for primary diagnosis after a study of about 2,000 surgical pathology cases.
• Digital slides support remote diagnosis, faster second opinions and tumor boards, and they reduce the physical transport of glass slides.
• AI tools built on digital slides can highlight suspected cancer, grade its severity and extract quantitative information that may help predict treatment response.
• Aegis Capital, a HealthTech & Longevity VC, invests in early-stage startups in Central and Eastern Europe working in diagnostics and AI-driven digital health.

What Is Digital Pathology in Clinical Practice?

Digital pathology is the acquisition, management and analysis of digitized glass slides, a definition used in a 2022 paper in the Journal of Pathology Informatics. A scanning device captures histology slides as digital slides, and pathologists then examine the tissue sections on a computer monitor instead of through the eyepiece of a microscope. The images can be stored, retrieved, annotated and shared through the same digital environment that holds other pathology information about the case.

Traditional microscopy ties each diagnosis to a physical object that exists in one place at a time. Digital slides remove that constraint: several specialists can review the same image, archived cases can be recalled in seconds, and software can measure features that the human eye can only estimate. Those capabilities explain the many benefits laboratories expect from the shift, and they also explain why implementing digital pathology requires more than buying a scanner.

From Glass Slides to Digital Slides

The analog workflow starts with tissue removed during a biopsy or surgery, which is fixed, embedded in paraffin, cut into thin tissue sections and stained on glass slides. A pathologist reviews those slides under a microscope and records the diagnosis in the laboratory information system. The glass slides are then filed in a physical archive, where retrieving an old case for comparison can take hours or days.

In a digital workflow, the same stained slides pass through a whole slide imaging scanner before review. The pathologist reads the digital image, while the glass slide stays safely stored as the physical original. The diagnostic process remains the same in substance, but every step that involves locating, transporting or comparing slides becomes faster.

Regulatory Milestones for Digital Pathology Systems

In April 2017, the FDA authorized the first whole slide imaging system for reviewing and interpreting digital surgical pathology slides, based on a clinical study of about 2,000 cases across multiple anatomic sites. The study found that diagnoses made on digital images were comparable to those made on glass slides. The College of American Pathologists noted that the FDA treated the authorized system as a closed unit, so replacing the scanner, viewing software or monitor with other components counted as off-label use.

The COVID-19 pandemic accelerated remote use. In March 2020, the US Centers for Medicare & Medicaid Services allowed pathologists to review slides remotely under certain conditions, and in April 2020 the FDA issued a temporary enforcement policy to expand the availability of devices for remote reviewing and reporting. Remote diagnosis moved from a pilot concept to a working option for many laboratories within weeks.

Key Components of Digital Pathology Systems

A complete digital pathology system combines three key components:

• Slide imaging devices: whole slide scanners that digitize glass slides at diagnostic resolution, often in large batches.
• Image management software: platforms that store, organize and display high resolution digital images alongside case data.
• Integration with laboratory information systems: links that connect each digital slide to the patient record, the order and the final report.

Each component has to perform reliably on its own, and the value appears only when all three work together in daily clinical practice.

Whole Slide Imaging (WSI) as a Virtual Microscope

Whole slide imaging (WSI) captures the entire tissue area of a slide, not just selected fields of view. Viewing software stores the image at several resolutions, so the pathologist can pan and zoom from a low-power overview to cellular detail, which recreates the experience of virtual microscopy with high resolution images on a screen. The same file can open on a workstation, a laptop or, for non-diagnostic review, a mobile device.

The files are large. According to the 2022 Journal of Pathology Informatics paper, an uncompressed 20x whole slide image of a large specimen such as a radical prostatectomy can exceed 10 gigapixels and require nearly 30 GB of storage, and scanning at the 40x magnification recommended for primary diagnosis can raise that to 100 GB. A single prostatectomy case can include from 20 to more than 100 slides, so storage planning is a core part of any rollout.

Did you know: Large pathology centers may produce more than 1 million digital slides and around 1 petabyte of uncompressed data per year, according to the same Journal of Pathology Informatics paper.

Image Management and Laboratory Information Systems

Image management software and related software applications turn digital slides into usable clinical records. It controls who can open a case, keeps images linked to the correct patient, and allows annotations and measurements to travel with the slide. Integration with laboratory information systems matters just as much, since a digital slide without its case context cannot support a diagnosis.

Poor integration is one of the most common reasons adoption slows after the scanners arrive. Laboratories that plan data flows, access rights and backup before go-live avoid running two parallel workflows for longer than necessary.

Remote Diagnosis and Collaboration

Digital slides make remote collaboration routine and offer significant benefits for smaller hospitals. A pathologist in a smaller hospital can send a difficult case to a subspecialist in minutes, which supports rapid referral of complex cases and enhancing access to expertise for patients in remote locations. Multidisciplinary tumor boards can review the actual whole slide images together, so oncologists and surgeons see the findings that shape their treatment decisions.

Education benefits in the same way. Digital training resources allow residents and students to study complete slides from rare cases, and teaching collections no longer degrade or break with repeated use.

Validation Before Clinical Use

Before a laboratory uses whole slide imaging for diagnosis, it has to confirm that its own system performs as expected. The 2021 guideline update from the College of American Pathologists recommends a validation set of at least 60 cases for one application, plus another 20 cases for additional applications such as immunohistochemistry, and a washout period of at least 2 weeks between reading the digital and glass versions of each case. If concordance between the two readings falls below 95%, the laboratory should investigate and remedy the cause.

The 95% benchmark reflects the evidence: across 33 studies in the College's systematic review, the weighted mean concordance between digital and glass diagnoses was 95.2%. Validation also depends on adequate training, since pathologists need time to become confident reading on screen.

Why the Pathology Workforce Needs Digital Tools

Demand for tissue diagnosis is rising faster than the number of specialists. A workforce census by the Royal College of Pathologists found that only 3% of UK histopathology departments had enough staff to meet clinical demand, and 78% had vacant consultant posts. In the College's 2025 census, only 5% of pathologists reported that all clinical demand was met within contracted hours, and 40% planned to reduce their working hours within five years.

The pressure will grow. The International Agency for Research on Cancer projects more than 35 million new cancer cases worldwide in 2050, a 77% increase from 2022, and most of those diagnoses will pass through a pathology laboratory. Digital pathology does not create new pathologists, but it lets the existing pathology workforce sustain diagnostic services by sharing cases across sites and focus their time on the slides that need expert judgment.

How Artificial Intelligence Supports Digital Pathology

Computational pathology applies machine learning and deep learning to digital slides, particularly image analysis of tissue structure and cell features. AI tools cannot work on glass slides, so digital pathology is the precondition for every algorithm in the field. Current AI models support pathologists in three main areas:

• Detection: flagging areas suspicious for cancer so they are less likely to be missed.
• Grading: estimating how aggressive a tumor is, which helps grade its severity in a consistent way.
• Quantification: automating cell counting and measuring biomarker expression to extract quantitative information from each slide.

Detecting and Grading Cancer

In September 2021, the FDA authorized the first AI-based software to help pathologists detect prostate cancer on digitized biopsy slides. In the supporting study, 16 pathologists reviewed 527 slide images, and the software improved cancer detection on individual slide images by 7.3% on average compared with unassisted reads, with no impact on the reading of benign slides.

Grading is harder to standardize, since pathologists often disagree on the growth patterns that define a Gleason grade. The PANDA challenge, reported in Nature Medicine in 2022, used 10,616 digitized prostate biopsies and showed that AI algorithms reached agreement with expert uropathologists of 0.862 on a US validation set and 0.868 on a European one. The authors concluded that AI-based Gleason grading should now be tested in prospective clinical trials.

Example: A laboratory scans a series of prostate biopsies, and the AI software marks two small areas on one slide as suspicious. The pathologist reviews those areas at high magnification, confirms a small focus of cancer, and grades it. The software speeds up the search, while the diagnosis remains the pathologist's decision.

Biomarker Detection and Treatment Response

Beyond detection, AI can read signals in routine stained slides that normally require separate tests. A 2019 study in Nature Medicine showed that deep learning can predict microsatellite instability directly from standard hematoxylin and eosin slides in gastrointestinal cancer. Microsatellite instability indicates which patients respond exceptionally well to immunotherapy, and not every patient is tested for it in clinical practice.

This line of research suggests that routine slides could possibly predict treatment response and help select patients for further testing. The same quantitative methods support biomarker detection in drug development and clinical trials, where consistent measurement across many sites matters. These applications still require prospective validation before they change standard patient care.

Barriers to Adopting Digital Pathology

The adoption of digital pathology demands investment before the benefits appear. Scanners, storage, network capacity and software licenses represent a substantial upfront cost, and financial constraints slow adoption most in hospitals with limited budgets. Integration with existing laboratory information systems can be complex, and data protection rules require secure storage and controlled access to patient images.

Workflow change is as demanding as the technology itself. Each laboratory must validate its system, train its staff and decide how long to keep glass and digital pathology workflows in parallel. Institutions that treat the rollout as a clinical transformation project, not an IT purchase, tend to move faster.

Tip: If you need a second opinion on a biopsy, ask the hospital whether your slides can be shared digitally with the consulting center, which can shorten the wait compared with shipping glass slides.

Why Diagnostic Innovation Needs Early-Stage Capital

Digital pathology is a disruptive technology built from hardware, software and clinical validation, and each layer requires funding long before revenue. AI tools in particular need large, well-annotated datasets, multi-site studies and regulatory work before they can support routine diagnoses. Industry leaders in the life sciences sector now use digital slides for research and trials, but many current and future applications, including AI tools that detect disease earlier, will come from early-stage companies. Laboratories embracing AI also need products that fit validated workflows, which is where specialized founders add value.

Aegis Capital is a HealthTech & Longevity VC that invests in early-stage startups from Central and Eastern Europe, with diagnostics and AI-driven digital health among its core investment areas. The fund has a capitalization of PLN 80 million and offers an initial ticket of up to PLN 3 million, with total funding of up to PLN 8 million per company across follow-on rounds.

FAQ

What is the difference between digital pathology and computational pathology?

Digital pathology covers scanning, storing, viewing and sharing slides as digital images. Computational pathology applies image analysis and AI to those images to detect, grade or measure disease.

Is a diagnosis on a digital slide as reliable as one on a glass slide?

The FDA authorized the first system for primary diagnosis after a study showed comparable results for digital and glass slides. Across 33 studies reviewed by the College of American Pathologists, the weighted mean concordance was 95.2%, and each laboratory must validate its own system.

How large is a whole slide image?

Size depends on the specimen and magnification. An uncompressed 20x image of a large specimen can require nearly 30 GB, and a 40x scan can reach 100 GB before compression.

Can AI replace pathologists?

No. Authorized AI tools support pathologists by flagging suspicious areas or measuring features, and the final diagnosis remains the pathologist's responsibility.

Where can pathologists find digital training resources?

Professional bodies such as the College of American Pathologists publish guidelines on validating whole slide imaging, and the Digital Pathology Association provides educational materials for the field. Many teaching hospitals also share digital slide collections for training.

‍