It’s been one day short of a year since Microsoft partnered with Adaptive Biotechnologies to create a blood test with the promise of diagnosing dozens of diseases at once. The idea is to take hold of the body’s own diagnostic powers by mapping the immune system’s response to toxins and foreign substances. To do that, the team needs lots of data and the ability to sort through it.
That’s where artificial intelligence comes in.
The companies announced Thursday that the AI system they built is up and running and will initially focus on diagnosing type 1 diabetes, celiac disease, ovarian cancer, pancreatic cancer and Lyme disease.
To generate the data, Adaptive and Microsoft placed an open call for collaborators to work toward their goal of sequencing immune data from 25,000 people with the five diseases. “Our AI systems are now ready. So please join us in decoding what ‘story’ the immune system is telling us,” wrote Peter Lee, Microsoft Corporate VP of AI and Research, on Twitter.
Our #AI systems are now ready. So please join us in decoding what "story" the immune system is telling us. Focusing today on type 1 diabetes, celiac disease, ovarian cancer, pancreatic cancer, and Lyme disease; more diseases in the future. https://t.co/wjTINSvJX1
— Peter Lee (@peteratmsr) January 3, 2019
The project will collect information on how T cells, white blood cells that are involved in immune response, bind to antigens. From there, the teams hope to create models that can accurately diagnose patients with existing diseases or genetic risk factors.
Artificial intelligence is at the center of an ongoing revolution in precision medicine, allowing researchers to make sense of data from healthcare devices. In addition to Microsoft, other major contenders in applying AI to health include Alphabet’s Verily Life Sciences, IBM Watson Health and Amazon Web Services.
“Even while we are still building capacity, we have already pushed the state-of-the-art in predicting antigen binding, refined our ability to diagnose a viral infection, and developed a method for using T-cell sequences to accurately estimate the genetic risk associated with many autoimmune diseases,” Jonathan Carlson, director of immunomics at Microsoft Healthcare NExT, wrote in a blog post. Microsoft’s Healthcare NExT program seeks out health industry partners to work with its research and AI teams.
Adaptive recently partnered with the University of Florida to profile thousands of patients at risk of or living with type I diabetes. It has also partnered with the Fred Hutchinson Cancer Research Center as well as groups at the University of Colorado and Virginia Mason.