Uncategorized

FDA Clears First AI Tool That Catches Pathology Slide Errors Before They Reach a Diagnosis

Leica Biosystems' newly FDA-cleared Aperio iQC DX caught up to 24% more slide defects than histotechnicians while cutting manual quality-review time by 69%, the company says.

FDA Clears First AI Tool That Catches Pathology Slide Errors Before They Reach a Diagnosis

Leica Biosystems said this month it has won U.S. Food and Drug Administration 510(k) clearance for Aperio iQC DX, the first AI-assisted quality control software cleared for clinical use in digital pathology, alongside clearances for two new whole-slide scanners, the Aperio GT 180 DX and an enhanced Aperio GT 450 DX. The announcement, detailed in a company release covered by Clinical Lab Products and News-Medical.net in mid-August, marks a shift in how AI is being deployed in cancer diagnosis: not to read the slide itself, but to make sure the slide is even readable in the first place.

What the software actually does

Aperio iQC DX runs while a glass slide is still sitting on the scanner, screening the resulting digital image for six common quality defects: air bubbles trapped under the coverslip, stray pen marks, clipped tissue that falls outside the scan area, missing tissue, out-of-focus regions, and image striping artifacts. Historically, catching these problems has depended on a histotechnician’s eye, often after a slide has already moved downstream toward a pathologist’s queue. Leica says its AI models flagged up to 24% more artifacts than histotechnicians identified on their own during clearance testing, while cutting the hands-on time needed for manual quality review by 69%.

Why a quality-control problem matters for cancer patients

Leica Biosystems, a Danaher Corporation subsidiary, says its instruments and reagents support close to 2 million cancer tests globally every week, making it one of the largest suppliers of anatomic pathology infrastructure in the world. A blurry or artifact-laden slide can force a re-cut and re-stain of tissue, adding days to a cancer diagnosis at a moment when patients are often waiting anxiously for a treatment plan. “Every improvement in the pathology workflow matters because patients and clinicians are waiting for answers that guide care,” said Naveen Chandra, vice president and general manager of digital pathology at Leica Biosystems, in the company’s announcement.

How digital pathology got here

Pathology has been one of the slower medical specialties to digitize, in part because whole-slide images are enormous files and labs have had to justify the capital cost of scanners against a manual microscope workflow that has worked, however imperfectly, for over a century. The FDA has been approving digital pathology hardware and AI software in a steady trickle since the first whole-slide imaging system was cleared in 2017. This year alone the agency has cleared AI tools from PathAI and Artera for direct diagnostic support; Leica’s clearance targets the workflow layer underneath those diagnostic tools, arguing that image quality has to be solved first or AI diagnostic software built on top of bad scans will simply inherit the errors.

A more skeptical read

Not everyone in pathology is convinced quality-control automation solves the field’s real bottleneck. Critics of piecemeal AI rollouts in pathology have long argued that the specialty’s biggest constraint is a shortage of pathologists themselves — the College of American Pathologists has flagged a shrinking workforce relative to rising biopsy volume — and that shaving minutes off slide-quality review does little to address that gap unless it is paired with AI that also speeds up diagnostic interpretation. There is also the practical question of integration: quality-control software only pays off if labs already run compatible Aperio scanners, meaning the benefit is initially confined to Leica’s own installed base rather than the pathology field broadly.

The competitive backdrop

The clearance lands in a crowded month for digital pathology AI. Artera won FDA clearance for ArteraAI Breast, a risk-stratification tool for breast cancer, in early May, while PathAI’s AISight Dx platform — cleared for primary diagnosis in 2025 — is now being deployed nationwide through an expanded partnership with Labcorp announced in February. Leica’s move to clear a quality-control layer suggests the next competitive battleground in pathology AI may not be diagnosis itself but the surrounding infrastructure: scanning, triage, and now automated quality assurance.

What comes next

Leica says the newly cleared Aperio GT 180 DX and enhanced GT 450 DX scanners will roll out to hospital and reference labs alongside the iQC DX software in the coming months, with the company positioning the trio as an integrated “scan-to-diagnosis” pipeline. Whether quality-control automation meaningfully shortens time-to-diagnosis at the patient level — rather than just the lab-workflow level — will depend on independent, peer-reviewed data that has not yet been published. For now, the clearance signals that FDA regulators are willing to treat pathology AI as a multi-layered stack, not just a single diagnostic algorithm, opening the door for more narrowly scoped AI clearances across the digital pathology pipeline.