How AI Learning Can Protect Patient Privacy and Still Offer Valuable Research

Healthcare organizations continue to make great strides in their use of artificial intelligence applications to improve patient care. But for AI systems to produce high-quality algorithms, they need large, diverse data sets.
Compiling these data sets can be a challenge because of regulatory and ethical obligations that restrict access to patient data. These obligations can lead chief medical informatics officers to adopt policies that forbid healthcare data from leaving an organization.
Reluctance to compile large data sets, driven by the increasing risks of financial penalties and…