Microsoft has developed a multimodal artificial intelligence model that converts routine pathology slides into spatial proteomics maps, which chart the distribution of proteins across tissue samples while preserving their physical arrangement.
Satya Nadella, Microsoft's chief executive, announced the work in a LinkedIn post, framing it as a way to reduce the time and cost of proteomic mapping and to widen access to cancer diagnostics and research.
The model, named GigaTIME, has been made available on Microsoft Foundry Labs and Hugging Face, the open-source AI platform, according to comments on the post.
Spatial proteomics is typically expensive and technically demanding, requiring specialist equipment and expertise that limits its use to well-resourced research and clinical settings.
Microsoft positions GigaTIME as a way to produce high-resolution protein maps from cheap, standard pathology slides that are already widely used in clinical practice.
Responses to the post highlighted that the approach is intended to augment rather than replace clinical expertise, and to extend access in settings where specialised pathology services are scarce.
Nadella described the result as an example of artificial intelligence shortening the path between scientific discovery and patient care.
Related reading
- Microsoft validates NVIDIA Vera Rubin NVL72 in the cloud
- Microsoft chief tests AI agent that completes multi-step tasks without manual handoffs
- Microsoft deploys Dragon Copilot at Manchester NHS trust to cut clinical paperwork
The LinkedIn post did not include a deployment timeline or details of clinical validation, leaving open questions about when and how the model might be used in a healthcare setting.
The announcement drew more than 3,000 reactions and 215 comments on a LinkedIn account with nearly 12 million followers.
The recap
- Satya Nadella announces AI converts pathology slides to spatial proteomics
- Post attracted 3,073 reactions and 215 comments on LinkedIn
- No timeline or implementation details were provided in the post