Biotechnology HPC Software Applications Institute (BHSAI)

A New Tool for Sleep-Deprived Soldiers to Optimize and Restore Alertness When It Matters Most

November 8, 2018   |  Download PDF

Sleep deprivation is common among military personnel. Roughly 40% sleep less than 5 hours a night, even though they need 7 to 8 hours to recover mental acuity. Yet, Warfighters often need to be vigilant while deprived of sleep. For example, in combat operations, they may face demanding schedules, ranging from a few days without sleep to many days of less than 5 hours a night. Inevitably, this hinders their cognitive performance and increases the risk of accidents.

Not surprisingly, caffeine is widely used in the military. (Around 85% of U.S. adults consume a daily amount equal to that in two cups of coffee.) As overdosing has negative consequences, individualized guidelines on when and how much caffeine to consume would be valuable. However, despite the boom in health-monitoring gadgets, no evidence-based tool has been developed to provide caffeine intake schedules that restore alertness in a timely and safe manner. Until now.

A team of scientists from TATRC’s Biotechnology High Performance Computing Software Applications Institute (BHSAI), led by Dr. Jaques Reifman, recently tackled this problem with collaborators at the Walter Reed Army Institute of Research. They asked whether an algorithm could be developed to estimate when to consume caffeine, and how much, to restore alertness at the desired time for the desired duration in sleep-deprived subjects without overdosing.

The ADMET predictions result page graphicA screenshot of 2B-Alert Web, which is accessible here. This tool uses a validated mathematical model to predict the effects of sleep loss and caffeine intake on alertness and an algorithm to find optimal caffeine schedules that maximize alertness at the desired time for the desired duration.  

The team built upon a comprehensive model of sleep, previously developed and validated extensively by Dr. Sridhar Ramakrishnan, a BHSAI staff scientist and co-author of the paper. The model predicts alertness, as measured by the gold-standard psychomotor vigilance test, based on a history of sleep schedules and caffeine intake events. Using this model, the BHSAI team first derived a mathematical function, which when minimized by optimizing its parameters, was expected to yield safe caffeine schedules that restored alertness in a temporally specific way.

One major roadblock stood in the way: conventional optimization algorithms took way too much time. Dr. Francisco Vital-Lopez, a BHSAI staff scientist and lead author of the study, ultimately came up with a workable solution. Drawing upon an earlier algorithm that could solve the well-known “traveling salesman problem,” and placing reasonable constraints on the allowable doses and time points, he found a way to simplify the process of searching for caffeine schedules that maximized alertness at a specified time, so that it required only a few seconds.

This breakthrough, together with 2B-Alert Web and 2B-Alert App—tools that BHSAI scientists have developed to predict the effects of sleep loss and caffeine at the group and individual levels, respectively—now sets the stage for finalizing an app that individual Warfighters can use to personally manage their caffeine and sleep schedules. Such an app, connected to a network, may also help commanders plan missions more effectively by quantitatively assessing Force status in real time and providing customized alerts and countermeasure recommendations.

Clearly, sleep deprivation is a concern for individuals worldwide. This is evident from the public interest in this new work. So, the day we can download an app—one that tells us how much coffee to drink, and when, to stay alert when we need to—may come around sooner than we think.

“This deals directly with Soldier readiness,” said Dr. Reifman. “If you could come to work, drink caffeine and have your mental acuity improved by 40% for four hours, wouldn’t you like that? That’s what we’re trying to do here.”

This article was published in the November 2018 issue of the TATRC Times.