ÌÇÐÄTV

The Tissue Image Analytics (TIA) Centre is pleased to announce the release of TIAToolbox v2.0.0, a major update to the opensource platform widely used for digital pathology and AI-driven tissue image analysis. TIAToolbox provides an end-to-end framework for working with whole slide images (WSIs), covering data loading, pre-processing, deep learning inference, postprocessing, and visualisation. It is designed for clinical, biomedical, and computational researchers and supports a broad range of pathology image analysis tasks.

A Re-Engineered Deep Learning Engine 

Version 2.0.0 introduces a completely redesigned deep learning engine built to meet the demands of modern pathology AI workflows. The new framework delivers ~20× faster processing on large WSIs with Improved scalability for gigapixel-scale images through unified, streamlined API across models and output formats

These improvements enable faster, more reproducible analysis of high-resolution pathology data and support the growing need for scalable AI pipelines.

Award Winning AI Models, Now Open Source 

The release includes the KongNet nucleus detection models, which achieved top performance in international challenges such as MONKEY, MIDOG, and PUMA. These state-of-the-art models are now openly available, supporting transparent and reproducible research across the computational pathology community.

Enhanced Support for Real World Workflows 

TIAToolbox v2.0.0 introduces several features designed to integrate smoothly into research workflows such as Standardised outputs (dict, Zarr, AnnotationStore, QuPath JSON), Multichannel / immunofluorescence image support, Interactive segmentation with SAM, & Cloud native WSI access for images stored on S3, Google Cloud, Azure, or HPC systems.

These additions make it easier to build scalable pipelines, explore large datasets, and view AI outputs in familiar tools such as QuPath and TIAViz.

A CommunityDriven Release 

TIAToolbox v2.0.0 reflects contributions from the TIA team, the wider user community, and collaborators across academia and industry. The TIA Centre acknowledges the support of the PRISM Lab and major funders including MRC, CRUK, British Council, NIHR, BigPicture (EU Commission), Innovate UK, and EPSRC.

Learn more: 
Full release notes and documentation are available at: