In a post on X, he noted, “Today in Cell, we published new research illustrating how AI can accelerate cancer discovery.”
“With GigaTIME, we can now simulate spatial proteomics from routine pathology slides, facilitating population-scale analysis of tumor microenvironments across numerous cancer types and hundreds of subtypes,” Nadella added.
GigaTIME was developed in partnership with Providence and the University of Washington.
Today in Cell, we published new research illustrating how AI can accelerate cancer discovery. With GigaTIME, we can now simulate spatial proteomics from routine pathology slides, facilitating population-scale analysis of tumor microenvironments across numerous cancer types and…
— Satya Nadella (@satyanadella) December 9, 2025
Nadella highlighted the importance of this advancement, emphasizing Microsoft’s dedication to enhancing AI in healthcare. “Created in collaboration with Providence and the University of Washington, we hope this initiative accelerates scientists’ journey from data to insights, revealing new connections between genetic mutations, immune response, and clinical outcomes, ultimately benefiting health globally,” he stated.
About GigaTIME
GigaTIME is an innovative AI tool that greatly accelerates the analysis of tumor microenvironments. It can computationally simulate the analysis of various proteins in mere seconds, allowing researchers to examine tens of thousands of scenarios simultaneously.
This tool could help pinpoint patients who will benefit from specific treatments, thereby enhancing patient outcomes. It is also anticipated to clarify why certain patients might not respond and how to address tumor resistance.
The model was trained on a dataset from Providence comprising 40 million cells with paired hematoxylin and eosin (H&E) and multiplex immunofluorescence (mIF) images across 21 protein channels. Utilizing GigaTIME with data from 14,256 cancer patients across 51 hospitals, researchers created a virtual population of roughly 300,000 mIF images covering 24 cancer types and 306 subtypes.
This virtual population uncovered 1,234 statistically significant correlations between mIF protein activity and key clinical features such as biomarkers, staging, and patient survival. These results were further validated independently on 10,200 Cancer Genome Atlas (TCGA) patients.
“To our knowledge, this represents the first population-scale study of tumor immune microenvironment (TIME) based on spatial proteomics. Such studies were previously unfeasible due to the lack of mIF data. By translating commonly used H&E pathology slides into high-resolution virtual mIF data, GigaTIME offers a groundbreaking research framework for investigating precision immuno-oncology through population-scale TIME analysis and discovery,” the company stated in a blog post.
Future Directions and Availability
GigaTIME has been made publicly accessible on Microsoft Foundry Labs and Hugging Face to promote research in precision oncology. The model can be adapted to incorporate additional spatial modalities and cell-state channels, and integrated into advanced multimodal frameworks to enable conversational image analysis.
