Flahy uses knowledge graphs to support AI-powered healthcare decisions

Singh said the knowledge layer is essential because it connects a company’s data layer to its model layer and tells the model which facts matter and what a specific data point implies. Flahy processes a large context of biological and clinical information spread across a large graph, he said, and that graph helps determine what decision tree or route a given person should take for better treatment decision-making.

Singh spoke with John Furrier for theCUBE + NYSE Wired: AI Luminaries, an interview series on theCUBE, SiliconANGLE Media’s livestreaming studio. The discussion focused on how knowledge graphs can connect patient information to support personalized healthcare decisions.

According to Singh, Flahy’s team has spent years building a graph-based information database and training its model to recognize relationships among data points. He described a hypothetical patient whose genetic mutation and cholesterol marker could affect how a new clinical finding is interpreted. “I’m looking for this one person [who] had a genetic mutation and this specific high cholesterol marker. If I have a new clinical signal, would that change how this person should be treated? That is a very different question. It’s a graphical question by design,” Singh said.

Flahy works with graph technology companies including Neo4j Inc., Singh said. One challenge is incorporating wearable-device readings into a graph alongside other health information so the system can interpret changes over time. “We have created our own proprietary engines,” he said. “If you look at clinical decision-making or the problem that I’m trying to solve, it is always a traversal problem. When it comes to longitudinal data, you have to find a way to put it in a specific graph. For me, that’s been a challenge which we are effectively trying to solve in terms of how to connect the dots across separate modalities.”

Singh said AI-powered healthcare requires a clear explanation of how information informs a decision. Flahy’s consumer offering, FlahyLife, combines biological and health information to guide next steps in prevention, early detection and treatment selection, according to the company. “We are working with leading clinical laboratories and health systems and trying to deploy our platform in terms of better clinical decision making and closing care gaps,” Singh said. “If you could simply connect the graph, you can reason in a correct way. You’re serving the right people for the right test at the right time.”