{"id":1120251,"date":"2025-01-17T01:43:13","date_gmt":"2025-01-17T09:43:13","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-story&p=1120251"},"modified":"2025-01-21T08:07:48","modified_gmt":"2025-01-21T16:07:48","slug":"wef_science","status":"publish","type":"msr-story","link":"https:\/\/www.microsoft.com\/en-us\/research\/story\/wef_science\/","title":{"rendered":"Microsoft at WEF"},"content":{"rendered":"\n
<\/span>
<\/div>
\n
\n
<\/div>\n\n\n\n

AI for Science: Learning the language of nature<\/h1>\n\n\n\n
<\/div>\n<\/div>\n<\/div><\/div>\n\n\n\n
\n
\n\t\n\t
\n\t\t
\n\t\t\t
<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n
\n
<\/div>\n\n\n\n
\n
<\/div>\n\n\n\n

At Microsoft, we believe that the ability of generative AI to learn the language of humans is equally matched by its ability to learn the language of nature.<\/h2>\n\n\n\n
\n
Health<\/a><\/div>\n\n\n\n
Discovery<\/a><\/div>\n\n\n\n
Earth<\/a><\/div>\n<\/div>\n\n\n\n
\n\n\n\n

Health<\/h2>\n\n\n\n

Diagnosis and treatment<\/h3>\n\n\n\n

Generative AI has revolutionized machines’ ability to understand human language and images, particularly in medicine, showing promise for improving patient outcomes and clinician experience.<\/p>\n\n\n\n

\n
<\/div>\n\n\n\n

Virchow<\/h4>\n\n\n\n

Microsoft Research, in collaboration with Paige (opens in new tab)<\/span><\/a>, a global leader in clinical AI applications for cancer, is advancing the state-of-the-art in computational foundation models. The first contribution of this collaboration is a model named Virchow. Virchow serves as a significant proof point for foundation models in pathology, as it demonstrates how a single model can be useful in detecting both common and rare cancers, fulfilling the promise of generalizable representations.<\/p>\n\n\n\n

\n
\n
\n\t
\n\t\t
\n\t\t\t\t\t\tModel<\/span>\n\t\t\tVirchow model on Azure AI Foundry<\/span> <\/span><\/a>\t\t\t\t\t<\/div>\n\t<\/article>\n<\/div>\n\n\n\n
\n\t
\n\t\t
\n\t\t\t\t\t\tNature MEDICINE Publication<\/span>\n\t\t\tA foundation model for clinical-grade computational pathology and rare cancers detection<\/span> <\/span><\/a>\t\t\t\t\t<\/div>\n\t<\/article>\n<\/div>\n<\/div>\n\n\n\n
\n
\n\t
\n\t\t
\n\t\t\t\t\t\tModel<\/span>\n\t\t\tVirchow 2 model on Azure AI Foundry<\/span> <\/span><\/a>\t\t\t\t\t<\/div>\n\t<\/article>\n<\/div>\n\n\n\n
\n\t
\n\t\t
\n\t\t\t\t\t\tBlog<\/span>\n\t\t\tLarge-scale pathology foundation models show promise on a variety of cancer-related tasks<\/span> <\/span><\/a>\t\t\t\t\t<\/div>\n\t<\/article>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n
\n
<\/div>\n\n\n\n

GigaPath<\/h4>\n\n\n\n

GigaPath (opens in new tab)<\/span><\/a> is a novel vision transformer that attains whole-slide modeling by leveraging dilated self-attention to keep computation tractable. In joint work with Providence Health System and the University of Washington, we have developed Prov-GigaPath, an open-access whole-slide pathology foundation model pretrained on more than one billion 256 X 256 pathology images tiles in more than 170,000 whole slides from real-world data at Providence.  All computation was conducted within Providence\u2019s private tenant, approved by Providence Institutional Review Board (IRB).<\/p>\n\n\n\n

\n
\n
<\/div>\n\n\n\n
\n