PNNL

Recent medical issues caused by natural disasters, acts of terrorism, and other human-made disasters have shifted public and political attention to the need for better situational awareness, predictive capabilities, and sharing of medical and public health information. Such events have resulted in increased demands on government resources for more timely, accurate, and relevant data to inform prevention, detection, and response activities. Currently, no electronic interoperable network exists that can provide real-time situational awareness of health-threat events, available medical resources, and potential future threats across the nation. Such a network would enable better-informed, data-driven decisions in critical times of need.

Approach

The MedINT platform enhances data-driven decision support through early warning and virtually assisted investigative capabilities. This flexible platform is founded on sound data science, software engineering, and analytic visualization tools that enable detection, tracking, characterization, and situational assessment of events that have the potential to, or are currently causing, harm to human, animal, or plant health, including threats to health system and food security.

MedINT combines multiple disparate data streams, including timely medical and open-source data, into a single access point, allowing end users to have real-time information at their fingertips. The MedINT Dashboard’s situational awareness and alerting capabilities ensure that users are aware of potential, ongoing, and anomalous threats, such as the next pandemic or possible shortage of necessary medical resources. Further, the dashboard is linked to MedINT Canvas, a virtually assisted investigation tool that lets users dig into the complete data and analytic repository. Here, users explore specific information of interest (e.g., the location of the threat or specific disease) or known knowledge about the threat to gain a deeper understanding and uncover hidden connections.

MedINT contains two analytic engines, OpenAAID and PhenoSage, that provide high-powered decision support to users through a One Health approach. OpenAAID uses a natural language processing text analytic pipeline to leverage open-source news, reports, and weather; scientific literature; and animal health data to aid in adaptable and automated insight-detection for digital medical intelligence. PhenoSage combines disparate medical records and systems to build a clinical knowledge graph using artificial intelligence to aid in anomaly detection and prediction. Ultimately, MedINT supports medical and public health decision-making by supplying the right information to the right person at the right time through cutting-edge data analytics, visualizations, and human-assisted user interfaces.

Benefits

MedINT, a comprehensive situational awareness platform powered by a virtual medical-intelligence assistant, will enable faster, better-informed medical investigations, decision-making, and public health response. The PNNL platform aims to provide global situational awareness of current and potential health threats to our nation; assist in early warning of present health risks through identification of threats at the patient level, medical facility, and incrementally higher geographic resolutions; and enhance data-driven investigations by harmonizing across multiple, heterogenous datasets to rapidly uncover key information. Coupling this real-time threat information and data-driven predictions with resource availability and on-the-ground status, decision-makers can make optimal real-time decisions that safeguard our health and security.