Protecting the healthcare enterprise from PHI breaches using streaming and NLP

Hospitals small and large are adopting cloud technologies, and many are in hybrid environments. These distributed environments pose challenges, none of which are more critical than the protection of protected health information (PHI). Jeff Zemerick explores how open source technologies can be used to identify and remove PHI from streaming text in an enterprise healthcare environment.

The healthcare technology domain has seen explosive growth in the past decade. Hospitals small and large are beginning to adopt cloud technologies, and many are in hybrid environments. These distributed environments pose unique challenges, none of which are arguably more critical than the protection of PHI. The federal regulations that govern PHI have strict requirements on how PHI is handled and have potentially severe consequences in the case PHI is left unprotected or exposed. These requirements place a significant burden on the operations teams who maintain the environments. These teams often have to manage multiple production, test, and development environments. Having a tool that can help protect PHI from being transmitted across environment boundaries can help the organization meet its regulatory obligations. Jeff Zemerick explores an architecture and the open source applications that implement it to identify and remove PHI from natural language text. He examines how streaming healthcare text can be ingested through Apache Kafka and processed by Apache Flink where natural language processing (NLP) methods identify and remove PHI. And you’ll see a demonstration of the deployment of the application to the cloud.