Paige, has partnered with Microsoft to unveil the second generation of its Virchow model, a powerful tool designed to improve cancer diagnosis and treatment. Known as Virchow2 and Virchow2G, these models are part of Paige’s broader effort to harness AI for better understanding and treating cancer.
Virchow2 and Virchow2G are based on a vast collection of over 3 million pathology slides, representing patients from 45 countries and more than 800 labs. This diverse dataset helps the models account for a wide range of factors, including different genders, races, ethnicities, and geographic regions. With 1.8 billion parameters, Virchow2G is currently the most extensive model ever created for pathology, significantly larger than previous versions.
Developed in collaboration with Microsoft, these models have benefited from advanced supercomputing resources, allowing them to achieve new levels of accuracy and efficiency in AI training. According to a recent report, this collaboration has set a new standard for AI in pathology.
Thomas Fuchs, the founder and chief scientist at Paige, highlighted the potential of these models to reshape cancer diagnosis. “We are just beginning to explore how these models can transform our understanding of cancer,” Fuchs said. The goal is to use these tools to provide more accurate diagnoses, tailor treatments more precisely, and ultimately improve patient care.
Razik Yousfi, Paige’s senior vice president of technology, emphasized how this new generation of the Virchow model is a step toward making precision medicine a reality. He noted that the models are not just about improving accuracy and speed but also about expanding the possibilities of what AI can do in the field of pathology.
In addition to aiding in cancer diagnosis, Paige is using its AI technology to support research and development in life sciences and pharmaceuticals. The company has developed several AI modules that help with therapeutic targeting, biomarker identification, and clinical trial design, potentially speeding up the development of new treatments.