Keywords: biomonitors, physiological endpoints, machine learning, predicitive algorithms
Physiological endpoints, such as sleep patterns, heart rates, respiration rate, and body temperature are important indicators of overall health status and are often disrupted when the body is not in homeostasis. Such disruptions are common with bacterial infections (anthrax, plague, tularemia, etc.), viral infections (Ebola, rift valley fever, Venezuelan Equine Encephalitis Virus, etc.), exposure to toxins (ricin, botulinum, Staphylococcal enterotoxin B, etc.), and high stress levels such as those encountered during routine military operations and/or deployments. Monitoring these baseline physiological parameters in real time could represent a useful method to assess current and future health status. To this end, this project’s ultimate aims are to develop an early warning system that monitors physiological endpoints using state-of-the-art commercial off-the-shelf (COTS) biomonitoring devices and correlates that data with actual health status and medical readiness. To date, numerous COTS devices have been evaluated using a small cohort of volunteers based on several operational criteria: PII security, performance, robustness, data security controls, and reliability in monitoring and recording physiological parameters of interest. In addition, statistical algorithms have been created using R to analyze subject time-series data. The algorithms monitor sleep, heart rate from inter-beat-intervals, and diurnal patterns, providing working baseline data. The presentation will discuss preliminary analysis of data for a number of subjects.