Get in touch →

Next-generation AI cancer tests,
in-house.

Slideflow Labs develops AI biomarker tests that run inside your hospital’s own infrastructure. Quantitative tumor biology read from the standard H&E slide your pathology team already prepares, returned to the treating physician in minutes.

Talk with our clinical team →Read the evidence summary
Runs whereInside your hospital’s infrastructure.
Reads fromThe H&E slide you already prepare.
ReturnsStructured biomarker output, in minutes.
Built and supported by
Platform

Precision starts with a deeper view of the slide.
For oncology testing, it’s Slideflow.

Our platform bridges the gap between digital histology and molecular intelligence. By leveraging advanced deep learning to decode complex tissue patterns, Slideflow transforms standard H&E slides into actionable insights, supporting more personalized approaches to cancer care.

How it works

Three steps that fit directly into your pathology workflow.

Slideflow has been tested and validated on real-world clinical workflows at leading cancer centers. Because it runs inside the digital pathology scanning infrastructure your team already operates, it fits how IT and pathology already work instead of adding a separate system to adopt and maintain.
Step 01

You scan the slide you already prepare.

The standard H&E section from routine surgical pathology. No additional stains, no additional specimen, no shipping.

Step 02

Models run securely on your infrastructure.

Slideflow runs as isolated, single-tenant containers in a dedicated cloud environment provisioned for your institution, not a shared multi-tenant API. Slides and case data stay inside that access-controlled environment over private networking, with no public data endpoints. The containerized deployment is built to pass your existing IT and security review.

Step 03

Results land in the systems your team already uses.

Quantitative biomarker output returns to the treating physician through your standard clinical reporting workflow, in minutes.

Selected publications

Built on peer-reviewed science.

  1. 01
    Slideflow: deep learning for digital histopathology with real-time whole-slide visualization.
    Dolezal JM, Kochanny S, Howard FM, et al.BMC Bioinformatics2024
  2. 02
    Uncertainty-informed deep learning models enable high-confidence predictions for digital histopathology.
    Dolezal JM, Srisuwananukorn A, et al.Nature Communications2022
  3. 03
    The impact of site-specific digital histology signatures on deep learning model accuracy and bias.
    Howard FM, Dolezal JM, et al.Nature Communications2021
  4. 04
    Generalization of a deep learning model for HER2 status prediction on H&E-stained whole-slide images.
    Howard FM, Villamar DM, He G, et al.npj Breast Cancer2024
  5. 05
    Machine learning–guided adjuvant treatment of head and neck cancer to improve survival.
    Howard FM, Kochanny S, Koshy M, Spiotto M, Pearson AT.JAMA Network Open2022

Beyond our own work, more than 100 peer-reviewed studies build on the open-source Slideflow framework.

See all research →Explore on Google Scholar →

Bring next-generation cancer tests in-house.

Contact us →contact@slideflow.ai
For researchers

The open framework.

The Slideflow framework remains open source. Researchers can build, evaluate, and publish on the same foundation we use to develop our clinical-grade tools. When a model is ready for the clinic, Slideflow Pro, our regulatory-grade deployment engine, takes it into practice.

Slideflow on GitHub

An end-to-end deep learning library for digital histopathology, developed by our team and used in peer-reviewed research worldwide. Open to contributors, citable in your work, and the foundation for everything we build.

MIT-licensed100+ publications built on Slideflow