Lessons learned from the COVID-19 pandemic underscore the need for more rapidly scalable, low cost, and broadly available approaches in responding to public health emergencies. Digital tools like smartphone or web-based applications present opportunities for new classes of medical countermeasures that could complement traditional approaches.
Through AI/ML algorithms, smartphone and computer applications can use images and audio captured with the phone or computer’s existing hardware to assess aspects of a person’s health. Health-related image and audio recordings contain much more information than the human eye or ear alone can process, but AI/ML algorithms could be used to process that information, detect differences and patterns among images and sounds, and identify disease signatures. These digital tools could empower individuals with actionable information about their infection status or acute skin condition and symptoms so that they can self-isolate and seek medical care if appropriate.
DRIVe is supporting research and development of these kinds of tools through the new Digital Medical Countermeasures program (formerly the Digital Tools for Diagnostics, Health Security and Pandemic Preparedness program) and has partnered with two companies developing algorithms and, eventually, smartphone apps for infectious disease symptom checking and detection.
Virufy is developing an app that uses a few seconds of cough audio, collected through a smartphone or computer microphone, to identify cough patterns indicative of a current COVID-19 infection. Virufy has demonstrated this capability based on an analysis of several thousand samples from international clinical networks and through crowdsourcing. With DRIVe’s support, Virufy will conduct additional clinical studies to refine and test their algorithm. Virufy will also initiate proof-of-concept work on AI algorithms to screen for other respiratory diseases based on cough audio.
VisualDx aims to create an AI/ML algorithm that analyzes user-provided skin lesion images together with a symptom questionnaire on a smartphone app. The algorithm can provide information on the indicated disease and other similarly presenting conditions. If successful, this project would demonstrate the power of AI/ML-based image analysis for easily and rapidly accessible symptom checking for potentially any pox-based virus such as smallpox or Mpox, and provide a digital health solution that could be modified easily in a future outbreak.
BARDA’s 2022-2026 Strategic Plan emphasizes the importance of preparing for future public health emergencies by collaborating with innovators to develop medical countermeasures that can help save lives. Cutting-edge, digital guidance tools that can pivot and be brought to scale in response to new threats could aid in public health response. These DRIVe projects, if successful, could demonstrate the power of these tools to rapidly respond to other threats such as influenza and smallpox.
These awards have been made under the Easy Broad Agency Announcement (EZ-BAA) area of interest (AOI) #9: Digital Medical Countermeasures.