Today, at the prestigious HIMSS conference, HistAI took the stage to announce a monumental leap forward in medical research and artificial intelligence: the launch of the SPIDER initiative — Supervised Pathology Image-DEscription Repository. This ambitious, year-long open-source project promises to redefine the landscape of pathology research by amassing an unprecedented dataset of over 50 million fully supervised images spanning 20 different organs and tissue types. With this initiative, we aim to empower researchers worldwide, unlock new insights into human tissues, and accelerate the development of AI-driven tools that will transform clinical diagnostics and drug development.

The SPIDER initiative is more than a project — it’s a global call to action. By training multi-class AI models capable of classifying over 400 distinct morphologies, we are laying the foundation for a shared resource that will fuel innovation, foster collaboration, and streamline workflows in pathology labs and pharmaceutical pipelines alike. This is a bold step toward a future where AI not only augments human expertise but also uncovers patterns and possibilities previously beyond our reach.

A Historic First Step: More than 4 Million Patches, 52 Morphologies, and Beyond

Kicking off this visionary endeavor, we are thrilled to release the first batch of the SPIDER project today — an extraordinary patch-based, fully supervised pathology dataset that stands as the largest of its kind. Derived from nearly 6,000 proprietary Whole Slide Images (WSIs), this initial release includes over 4 million patches representing three critical human tissues: Skin, Colorectal, and Thorax. Accompanying this dataset are three cutting-edge patch-classification AI models — one for each organ — capable of accurately identifying 52 unique morphologies with remarkable precision.

This release is a testament to months of tireless work and a commitment to excellence. Each image has been meticulously annotated and supervised, ensuring the highest quality data for researchers and developers. Whether it’s advancing diagnostic accuracy or uncovering novel therapeutic targets, this dataset is poised to become a cornerstone resource for the global scientific community.

What’s Next: Breast Tissue and Beyond

The momentum doesn’t stop here. Over the next two weeks, we will expand the SPIDER repository with a similarly robust dataset and AI model for Breast tissue, covering 25 morphologies. This addition will further broaden the scope of the project, offering researchers yet another powerful tool to explore and innovate.

Looking ahead, the SPIDER initiative will continue to grow, with plans to incorporate data from all 20 targeted organs and tissue types throughout the year. By making these datasets and models openly accessible, we are inviting researchers, clinicians, and technologists from every corner of the globe to join us in this transformative journey.

A Vision for the Future

The SPIDER initiative is more than just a collection of images and algorithms — it’s a catalyst for progress. By democratizing access to high-quality, supervised pathology data, we aim to accelerate the development of next-generation AI algorithms and products that will redefine how diseases are diagnosed and treated. From streamlining clinical workflows to expediting drug discovery, the ripple effects of this project will be felt across healthcare for years to come.

We invite the global research community to explore this first release, contribute to the growing repository, and build upon this foundation. Together, we can unlock new knowledge, push the boundaries of computational pathology, and ultimately improve lives worldwide.

For more information about the SPIDER initiative, to access the datasets, or to collaborate with us, visit our dedicated GitHub page. The future of pathology starts now — and it’s open to all.

Stay tuned for updates as we continue to roll out this groundbreaking project over the coming months!