(Released 7 June 2018) Biothreats — harmful pathogens that are either naturally or deliberately released — pose a risk to national security and public health, and identifying new detection methods is a top priority for the Department of Homeland Security (DHS) Science and Technology Directorate (S&T). Biothreats are hard to immediately identify, but with new technologies and data sources, such as the wealth of open data generated by “smarter” cities, emergency managers may be able to detect and respond to an emerging problem more quickly.
DHS S&T knew it would need to go beyond its four walls to explore this opportunity space. In October, S&T collaborated with the Office of Health Affairs National Biosurveillance Integration Center (NBIC) to launch the Hidden Signals Challenge. The $300,000 prize competition called for the design of an early warning system to keep our communities safe by using existing data sources to uncover emerging biothreats.
Moving from concepts to system designs
The Challenge attracted concepts from dozens of data science experts across the United States. A 17-person review panel of experts in data, health, and public safety scored the submissions before passing them on to the panel of judges with expertise in bioinformatics, biological defense, epidemiology, and emergency management.
In February 2018, DHS announced the five finalist teams, whose concepts employ a variety of data sources, machine learning approaches, and analytical models. Each finalist was awarded $20,000 as seed funding to enter the second stage of the Challenge, an eight-week Virtual Accelerator, where they further developed their concepts into detailed system designs.
The finalists each had different strengths and areas of expertise, so the Virtual Accelerator was designed to expose them to new perspectives that would shore up any knowledge gaps and inform their system designs. The program was split into four modules, each consisting of virtual guest speakers, reading lists, and guided field exercises on topics ranging from data science to design thinking and city operations.
The series kicked off with a session led by Aaron Firoved, Director at NBIC, and Tom McGinn, Senior Health Advisor at NBIC who provided a briefing on NBIC and its goals, discussing relevant biothreat case studies and goals for future detection technology. After this deep dive into federal biodefense systems, the teams went deep on data science with a module that included a Q&A with Graham Dodge, CEO of Sickweather on social listening, semantic analysis, and modeling for real-time analytics.
Deployment and testing was emphasized throughout the series. Kimberly Lucas, Director of Civic Research for the City of Boston said, “There is probably no such thing as too early for testing, but there is definitely such a thing as too late.” Chelsea Mauldin, Executive Director of the Public Policy Lab similarly endorsed testing early and often, reminding teams that “The sooner you get something in the hands of a user, the sooner you’ll figure out how broken it is.”
Throughout the Accelerator, the finalists also received one-on-one mentorship from seven experts, who offered insight into areas such as user research, design thinking, and system implementation planning to inform the finalists’ next round of submissions.
The teams noted how the experience provided new perspectives that informed critical design decisions. “These conversations strengthened our understanding of the complexity of how local and national organizations work together to detect and respond to emerging bio-threats. This understanding has helped us anticipate the needs of our various end users and consider how we can best scale our system to a national level,” said Dr. John Brownstein, Director of the Computational Epidemiology Lab at Boston Children’s Hospital from the Pandemic Pulse team.
“Through the Hidden Signals Virtual Accelerator, we’ve come to realize we need our platform to be flexible enough to allow for zooming in and panning out at any given moment, as this will increase understanding and therefore trust among our end-users,” said Daniel B. Neill, Director of the Event and Pattern Detection Laboratory at Carnegie Mellon University from the Pre Syndromic Surveillance team said. “The end-user buy-in is a critical step in development, as an under-utilized platform would have limited public health impact, particularly in times of emergencies.”
The winner and runner-up
Following the Virtual Accelerator, teams submitted their final white papers for evaluation by the judges. Last week, DHS announced the grand prize winner and runner-up who received prize money as rewards for their noteworthy progress towards deployable systems. Pandemic Pulse was named the winner of the $150,000 grand prize and Pre-syndromic Surveillance was named runner-up and received a $50,000 prize.
Pandemic Pulse was created by the Computational Epidemiology Lab at Boston Children’s Hospital. The team is widely known in public health circles for its work on infectious disease monitoring tools, leveraging Flu Near You and HealthMap. Their system provides an intuitive dashboard that overlays these two data sources with Twitter and Google Search data to detect biothreat signals. The tool then uses a tiered evaluation method to filter data based on pathogen category, information source, and transmission mode.
The runner-up, Pre-syndromic Surveillance by Daniel B. Neill and Mallory Nobles, uses semantic analysis to integrate emergency department chief complaints with data from health clinics and social media to discover outbreaks that do not correspond with known illnesses. With a pilot already in market, the team has developed a working prototype with New York City’s Department of Health and Mental Hygiene and other city agencies.
"By exploring these untapped data sources we aim to improve how city-level operators make important public safety decisions," said William N. Bryan, DHS Senior Official Performing the Duties of Under Secretary for Science and Technology. "The grand prize winner and runner-up have strong system designs that harness streams of information in a manner that could allow us to identify an emerging problem faster."
Last week’s announcement concludes the second stage of the Challenge. Building off of these successes, DHS S&T, local operators, and the participating teams will continue their quest toward deploying an early warning system to uncover emerging biothreats.
Released by the U.S. Department of Homeland Security Science and Technology Directorate. Click here for source.