San Francisco-based vendor Flow Health has formed a five-year partnership with the U.S. Department of Veterans Affairs (VA) to build a medical knowledge graph with deep learning to inform medical decision-making and train artificial intelligence (AI) to personalize care plans.
Through the collaboration, Flow Health will build a knowledge graph of medicine and genomics from over 30 petabytes of longitudinal clinical data drawn from VA records on 22 million veterans spanning over 20 years. All patient information will be de-identified during analysis to protect privacy.
According to a press release, the partnership’s fundamental objective is to understand the common elements that make certain people susceptible to particular diseases, pinpointing effective treatments and identifying possible side effects in order to inform care decisions. The collaboration will integrate large volumes of data and aims to discover relationships between genomes and phenotypes, in order to identify disease risk, make more precise diagnoses and to suggest individualized treatments.
“Developing artificial intelligence which can automatically identify the best diagnostic and treatment pathways will assist clinicians in delivering precision medicine to every veteran," Robert Rowley, M.D., chief medical officer at Flow Health, said in a statement. “To build artificial intelligence you need huge amounts of data to feed deep learning models. This is why this partnership between the VA and Flow Health is a watershed moment for deep learning in healthcare.”
