For most of the history of pathology, a cancer diagnosis depended on a physical glass slide, a microscope, and a single pathologist's trained eye. That fundamental process is now being digitized — and once a tissue sample becomes a high-resolution digital image rather than a physical slide, it becomes something AI can analyze, quantify, and cross-reference at a scale no individual reader could match.
What Whole-Slide Imaging Actually Changes
Whole-slide imaging (WSI) scanners convert glass slides into gigapixel digital files — a single slide can generate an image with over 100,000 x 100,000 pixels. Once digitized, that slide can be shared instantly with subspecialist pathologists anywhere in the world for a second opinion, archived permanently without physical degradation, and — critically — analyzed by computer vision algorithms trained to recognize specific cellular patterns associated with malignancy, grade, and molecular subtype.
The FDA has now cleared multiple AI-based digital pathology systems for primary diagnostic use in prostate cancer, where algorithms can identify and grade cancerous tissue with concordance rates matching or exceeding inter-pathologist agreement — a meaningful benchmark given that pathologist-to-pathologist disagreement on Gleason grading has historically run as high as 30-40% in borderline cases.
Quantifying What Used to Be Subjective
Beyond simple cancer detection, AI models can now quantify features that pathologists traditionally assessed qualitatively: mitotic count, tumor-infiltrating lymphocyte density, and HER2 or PD-L1 expression scoring on immunohistochemistry stains. This matters clinically because these scores directly influence treatment decisions — PD-L1 expression level, for instance, determines eligibility for certain immunotherapy regimens. Consistent, quantitative scoring reduces the variability that has historically made borderline cases a source of inter-lab disagreement.
The Turnaround Time Problem AI Is Solving
Anatomic pathology has faced a persistent workforce shortage — the U.S. pathologist workforce has declined even as biopsy volume has climbed with an aging population and expanded cancer screening. AI-assisted triage helps by pre-screening slides and prioritizing likely-malignant cases to the top of a pathologist's queue, while routing clearly benign findings for faster sign-off. Several large health systems report meaningful reductions in average diagnostic turnaround time after implementing AI triage — a difference that matters enormously to a patient waiting for a biopsy result.
Barriers to Wider Adoption
Despite the clinical promise, digital pathology adoption in the U.S. lags notably behind countries like the Netherlands, where several national health systems have gone fully digital. The primary barriers are economic rather than technical: WSI scanners cost $150,000–$500,000+ per unit, digital storage for gigapixel files at scale is expensive, and reimbursement codes for AI-assisted pathology review remain limited compared to traditional glass-slide sign-off. Smaller independent labs, in particular, have struggled to justify the capital investment without a clear reimbursement pathway.
Molecular Pathology Convergence
The more transformative long-term shift is the convergence of morphologic and molecular data. AI models are increasingly being trained to predict a tumor's likely genomic mutation profile directly from its visual appearance on a standard H&E-stained slide — a capability that, if validated at scale, could flag candidates for confirmatory genomic testing far more cheaply and quickly than sequencing every specimen.
Conclusion
Digital and AI-assisted pathology represents one of the clearer near-term wins in clinical AI: faster turnaround, more consistent grading, and instant access to subspecialist review. The technology is proven; the remaining barrier is largely financial. As reimbursement structures catch up to the clinical evidence, digital pathology adoption is likely to accelerate meaningfully over the next several years. Laboratories building out digital pathology capacity can find relevant lab supplies and diagnostic equipment in our catalog.
