Last updated: October 2026
The leading AI tools for dermatopathology in 2026 are Primaa’s Cleo Skin, PathAI’s PathAssist Derm and Proscia’s DermAI, with research models like HistoGPT showing where the field is headed. They help dermatopathologists triage skin biopsies, classify common lesions, take measurements such as margins and lesion dimensions, and flag features like mitoses and perineural invasion.
Dermatopathology is a natural fit for AI. Case volumes are high, many cases are routine, and a small share of difficult melanocytic lesions demand the most time and expertise. Unlike prostate pathology, though, dermatopathology AI is still early: as of October 2026, we couldn’t find an FDA-cleared AI tool for diagnosing skin biopsies. Below, we compare the leading tools, explain what each one does today and cover how labs are using them.
In this guide
- What dermatopathology AI does
- Comparison at a glance
- The top AI tools for dermatopathology
- How to choose a dermatopathology AI tool
- Dermatopathology AI in your workflow
- Frequently asked questions
What Dermatopathology AI Does
Dermatopathology AI analyzes digital images of skin biopsy and excision slides. It is different from dermatology AI, which analyzes photos of the skin itself. Today’s dermatopathology tools generally help with:
- Triage and prioritization: sorting the worklist so likely melanomas and other high-risk cases are read first.
- Lesion classification: suggesting likely categories, such as melanoma, nevus, basal cell carcinoma or squamous cell carcinoma, for the pathologist to confirm.
- Measurements: automating lesion dimensions and surgical margins, which reduces manual ruler work.
- Feature detection: flagging mitoses, mitotic hotspots and perineural invasion.
- Workflow help: orienting slides automatically and, in research, drafting the first version of a report.
In every case, the dermatopathologist reviews the AI output and makes the diagnosis.
Comparison at a Glance
| Tool | Developer | What it does | Regulatory status | Works in Lumea’s viewer |
|---|---|---|---|---|
| Cleo Skin | Primaa | Triage and classification of four common lesions, margins and dimensions, mitoses, perineural invasion | CE-IVDR certified (Europe); not FDA-cleared | Yes |
| DermAI | Proscia | Classifies skin biopsies into pre-diagnostic categories for triage and quality review | Research use only | No |
| HistoGPT | Helmholtz Munich and Technical University of Munich | Drafts dermatopathology reports from whole slide images | Research model | No |
| PathAssist Derm | PathAI | Orients slides, prioritizes the worklist, identifies 17 lesion types and measures lesions | Research use only; FDA Breakthrough Device designation (2026) | No |
Status as of October 2026. Check current regulatory status and intended use with each developer.
The Top AI Tools for Dermatopathology
1. Primaa: Cleo Skin
Founded in Paris in 2018, Primaa builds AI for breast pathology and dermatopathology. Cleo Skin triages the four most common skin lesions, melanoma, nevus, basal cell carcinoma and squamous cell carcinoma, so high-risk cases rise to the top of the worklist. It also detects mitotic figures, mitotic hotspots and perineural invasion, and takes sub-millimeter measurements of lesion dimensions and surgical margins. In 2026, Cleo Skin received CE marking under the EU’s In Vitro Diagnostic Regulation (IVDR), which Primaa says makes it the world’s first CE-IVDR certified AI tool for dermatopathology. It is not FDA-cleared, so U.S. labs use it for research or validate it themselves as a laboratory-developed test. It is available on several digital pathology platforms, including Lumea’s Viewer+.
2. PathAI: PathAssist Derm
PathAI launched PathAssist Derm in February 2025 on its AISight image management system. The tool automatically orients each tissue section so the slide is ready to read, prioritizes the worklist based on predicted lesion type, and can identify 17 skin lesion entities, from common ones like actinic keratoses and basal cell carcinomas to rarer ones like melanomas. It also measures lesions and suggests relevant ancillary stains. In March 2026, the FDA granted PathAssist Derm Breakthrough Device designation, which is meant to speed its review; it is still for research use only, not clinical diagnosis. In May 2026, Roche agreed to acquire PathAI.
3. Proscia: DermAI
Proscia launched DermAI in 2019 as one of the first commercial AI applications for dermatopathology. Running on Proscia’s Concentriq platform, it uses deep learning to classify skin biopsies into pre-diagnostic categories, which labs can use to sort and prioritize cases and as an automated second layer of review. It was validated on more than 20,000 biopsy slides from academic and commercial dermatopathology labs. Proscia said at launch that it planned to seek FDA clearance, but DermAI remains available for research use only.
4. Helmholtz Munich and TUM: HistoGPT
While most dermatopathology AI detects or measures, HistoGPT points to what comes next. Developed by researchers at Helmholtz Munich and the Technical University of Munich with collaborators including Mayo Clinic, it is a vision-language model that writes a first draft of the pathology report directly from whole slide images. It was trained on 15,129 slides from 6,705 dermatology patients and published in Nature Communications in 2025. HistoGPT is a research model, not a commercial product, but report drafting is likely to be one of the biggest time savers as these tools mature.
Looking for AI beyond dermatopathology? See our list of the top 10 AI companies for digital pathology.
How to Choose a Dermatopathology AI Tool
- Start with your bottleneck. If your dermatopathologists lose time to sorting routine cases, a triage tool helps most. If margins and measurements slow down reporting, look for measurement features.
- Check the regulatory path. Without an FDA-cleared option, U.S. labs using dermatopathology AI clinically validate it themselves as a laboratory-developed test. Plan for that work up front.
- Test it on your cases. Lesion mix varies widely between practices. Ask for performance data on cases like yours and run your own before you commit.
- Keep it in the viewer. AI that appears alongside the slide you’re signing out gets used. AI in a separate application usually doesn’t.
For a full plan, see How to Implement Digital Pathology.
Dermatopathology AI in Your Workflow
Lumea’s digital pathology platform for dermatology brings slide review, AI, clinician-specific reporting, molecular ordering and billing into one workspace. Primaa’s Cleo Skin runs directly inside Lumea’s FDA-cleared Viewer+, so dermatopathologists see AI triage, measurements and biomarker detection on the same screen where they sign out, without changing their LIS or scanner. See how Utah Valley Dermatology brought its dermatopathology workflow onto Lumea.
Frequently Asked Questions
Is there FDA-approved AI for dermatopathology?
As of October 2026, we couldn’t find an FDA-cleared AI tool for diagnosing skin biopsies. PathAI’s PathAssist Derm received FDA Breakthrough Device designation in 2026 but is still research use only, and Primaa’s Cleo Skin is CE-IVDR certified in Europe. U.S. labs that want to use dermatopathology AI clinically validate it themselves as a laboratory-developed test.
Can AI diagnose melanoma?
AI can flag and prioritize likely melanomas and other high-risk lesions, but the dermatopathologist makes the diagnosis. Distinguishing melanoma from benign melanocytic lesions is one of the hardest calls in pathology: a large study published in The BMJ in 2017 found that, for the most challenging melanocytic lesions, pathologists agreed with a consensus diagnosis less than half the time. That’s where AI triage and a second look can help most.
What’s the difference between dermatology AI and dermatopathology AI?
Dermatology AI analyzes photos of the skin, often to help decide whether a lesion should be biopsied. Dermatopathology AI analyzes the microscope slides from that biopsy to support the diagnosis.
How do dermatopathology labs get started with AI?
Most labs first go digital for primary diagnosis, then add AI for one task, such as triage or measurements, and validate it on their own cases. Choosing a platform that runs third-party AI inside the viewer makes it easier to add tools as the field matures.
See Dermatopathology AI in Your Workflow
Want to see Cleo Skin and Lumea’s Viewer+ on your own cases? Request a demo.
