GigaTIME AI tool advances cancer research
Microsoft Research’s GigaTIME converts routine pathology slides into detailed digital maps to study tumour microenvironments at scale.
GigaTIME, an AI tool from Microsoft Research, lets researchers study the neighborhood around tumours at a larger scale than before.
Microsoft Research said in a blog post that GigaTIME uses multimodal models to turn routine pathology slides into detailed digital maps of protein activation, revealing how immune cells interact with cancerous tumours.
The blog noted that producing such maps traditionally takes days and costs thousands of dollars per sample; GigaTIME can simulate analysis across dozens of proteins in seconds, enabling researchers to compute tens of thousands of scenarios at once.
Developed in collaboration with Providence and the University of Washington, GigaTIME was trained on research-approved data from more than 14,000 patients across 51 hospitals and over 1,000 clinics, the blog said. The tool is publicly available as open-source on Microsoft Foundry Labs and Hugging Face.
Microsoft said GigaTIME could help identify which patients might benefit from specific treatments and illuminate reasons for treatment resistance.
The Recap
- GigaTIME maps tumour microenvironments from routine pathology slides.
- Trained on data from more than 14,000 patients in research.
- The tool is available open-source on Foundry Labs and Hugging Face.
