Introducing the Lagotto image search engine

A fundamental shift in computational pathology.

Whereas AI involving classification and segmentation methods has obvious benefits for image analysis, image search represents a fundamental shift in computational pathology. Using image search, pathologists can match the pathology of new patients with already diagnosed and curated cases, offering a new approach to improve diagnosis through visual inspection of similar cases and computational majority vote for consensus building.

Lagotto™ is a patented, content-based image retrieval solution that enables pathologists, researchers and educators to instantly find similar digital slides within their institution (or around the world) and access the knowledge in corresponding diagnostic reports, patient outcomes, and metadata.

Lagotto connects pathologists, researchers and educators to vast, collective knowledge.

With Lagotto, users can find similar cases and read multiple pathology reports from subspecialty experts in real-time. Through this “virtual peer review,” pathologists, researchers, and educators can search for similar images in their institution or across the globe, and then utilize the previously diagnosed pathology reports from trusted colleagues.

Uncovering new connections between pathology, genomics, treatments plans and patient outcomes.

The key to the Lagotto platform is its ability to easily find similar images in large archives and then access the associated diagnostic data. The pathology report is just the beginning, as patient records also include treatment plans, patient outcomes, and even data from genomics.

By connecting these rich datasets, Lagotto sets the foundation for new applications, such as predictive algorithms for personalized treatment plans or improved drug discovery.

It all begins with indexing whole slide images.

Whole slide images are large, often 1 gigabyte in size or more. Processing these images at full resolution is too slow and requires unmanageable amounts of additional storage.

The core of the Lagotto image search platform is a patented, artificial intelligence-based technique that indexes large, whole slide image archives. Millions of images are each represented by compact, super-efficient barcodes. Once indexed, the entire whole slide archive can be instantly searched by simply comparing barcodes. This enables Lagotto to search 100 million patches or 100,000 whole images per second.

Universal workflow compatibility

Index images from Huron or 3rd party scanners

This starts with the ability to index whole slide images from both Huron TissueScope scanners and virtually any 3rd party scanner.

Use your favorite image viewer

With Lagotto’s robust API, we can overlay image search functionality onto 3rd party image management/PACs systems, allowing pathologists to take full advantage of image search while using the existing viewing tools they already love.

On premise, on the cloud or hybrid

Lagotto can be deployed on premise, in the cloud, or in a hybrid storage environment, offering further flexibility.

Reference Architecture

Image Search Resources

Image Search Validation in Nature Digital Medicine

Huron Granted US Patent for Barcoding Images

Image Search in Elsevier’s Medical Image Analysis

Searching is Intelligence

Huron AI Advisor Recognized in Power List 100

Image Search PowerChat with Huron CEO