A digital pathology collaboration between Microsoft, the University of Washington and Providence Health Network aims to overcome a few obstacles to fully implementing artificial intelligence in cancer diagnostics—and, in some cases, through sheer scale.

Through a blog post, Microsoft introduced "GigaPath," 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 pre-trained on more than one billion 256 X 256 pathology image tiles in more than 170,000 whole slides from real-world data at Providence. All computation was conducted within Providence's private tenant, approved by Providence Institutional Review Board (IRB)", stated the blog.

According to the company, this is the first whole-slide foundation model for digital pathology, with large-scale pretraining on real-world data. Prov-GigaPath attains state-of-the-art performance on standard cancer classification and pathomics tasks, as well as vision-language tasks. This demonstrates the importance of whole-slide modeling on large-scale real-world data and opens new possibilities to advance patient care and accelerate clinical discovery.

GenAI and biomedicine

The launch of the new project was announced at a time when the confluence of digital transformation in biomedicine and the current generative AI revolution creates an unprecedented opportunity for drastically accelerating progress in precision health. Digital Pathology is an exciting frontier. In cancer care, whole-slide imaging has become routinely available, transforming a tumour tissue microscopy slide into a high-resolution digital image. Such whole-slide images contain key information for deciphering the tumour microenvironment, which is critical for precision immunotherapy.

Microsoft states that the present scenario is exciting, tempered by the reality that digital pathology poses unique computational challenges, as a standard gigapixel slide maybe thousands of times larger than typical natural images in both width and length. Research before in digital pathology often ignores the intricate interdependencies across image tiles in each slide thus missing important slide-level context for key applications such as modeling the tumor microenvironment. 

Microsoft Blog