Skip to content

RationAI/xopat

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1,589 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

xOpat — Explainable Open Pathology Analysis Tool

A web-based whole-slide imaging framework — not a viewer, but the viewer you configure to your use-case.

The xOpat Viewer

⚠️ This is xOpat v3 — an alpha release. The older but stable v2 line lives on the archive/v2 branch. If you experience anny issues, please, let us know.

What is xOpat?

Whole-slide imaging produces massive, complex data — scans and derived outputs in both vector and raster form — that are hard to work with efficiently. Patent restrictions and competing standards have fragmented the landscape into many ad-hoc open-source and proprietary tools. That diversity creates friction: clinicians face legacy systems and economic constraints; researchers juggle incompatible platforms and unwieldy data scales. And the viewer — the direct interface to all of it — usually bakes in one platform's API, forcing institution-wide, multi-petabyte decisions on the basis of viewer features that are often orthogonal to actual viewing needs. The result is cascading lock-in that can be impossible, or economically unjustifiable, to undo.

xOpat takes a different stance: it is not a viewer, it is a viewing framework. Rather than hardwiring an API or a deployment model, it adapts through modules, plugins, and deep configuration flexibility — letting you assemble a tailored interface for clinical, research, or educational use without committing your infrastructure to anyone's platform. The OpenSeadragon feature driver, and author of the Flex Render renderer, it is server-agnostic, modular, and extensible solution.

We run it across very different environments — closed-source infrastructures, custom data backends, ad-hoc configurations, Jupyter Notebook & Google Colab workflows, local desktop viewing, Kubernetes-scale deployments, and more.

What it gives you

  • 🔌 No hardwired backend. Connect to any WSI image server that speaks a protocol xOpat can resolve — add support for your own with a single file.
  • 🗂️ Vector and raster overlays. Photoshop-style visualization layers (WebGL shaders) and versatile annotations over the same slide.
  • 🧩 Plugins & modules. OIDC auth, AI chat assistants, visualization storytelling, tutorials, ICC profiles, action shortcuts, and more — load only what a deployment needs.
  • ⚙️ Static and runtime configuration. Shape the whole experience from env.json down to per-session URLs.
  • 🖥️ Run it your way. Node.js server, PHP server, or server-less — compiled once and served as static files.

What xOpat is not

It is not an all-in-one product. xOpat does not read WSI formats itself and does not run your AI — but it connects to the servers and services that do. If an image server can read your slides, xOpat can show them; AI results arrive as raster images (like any slide) or as vector graphics. Add a plugin to wire up whatever else you need.

Get started

You want to… Go to
Try it locally in one command Quick Start
Host it for real (Node / PHP / static) Deployment overview
Configure a deployment — secrets, proxy, IO Administration & Integration
Learn the terms Glossary
See what ships Plugins · Modules
Build a plugin or module plugins/README.md · modules/README.md
Understand the core Core architecture

📚 Full documentation: https://xopat.org  · 

RationAI Logo

Sponsors

We are grateful for the development and financial contributions supporting xOpat and the OpenSeadragon project.

BBMRI ERIC Logo

About

Histopatholical visualization toolkit based on OpenSeaDragon.

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors