Picture the scene. It’s the early hours of the morning. You’ve been out partying with friends, and somehow the hours have just slipped by. You try and drive home as best you can, sticking to the speed limit and staying in your lane, but your reflexes aren’t what they should be. Your vision is blurry, your reactions slowed. Suddenly the lights of a cop car appear in your rearview mirror. You pull over to the side of the road and wind down your window. The cop walks over. “Excuse me,” they say. “Have you been sleeping properly?”
OK, so that’s probably not the first question they might have in this scenario, but perhaps it should be. While drunk driving is undoubtedly a big problem, the impact of sleep loss on a person can be comparable to the cognitive impairment caused by alcohol consumption. Fortunately, there may soon be a way to easily and objectively measure sleep loss — much the same way that a breathalyzer provides a non-subjective way to ascertain a person’s levels of blood alcohol.
At the Sleep Research Center at the U.K.’s University of Surrey, researchers have used a machine learning algorithm to help develop a blood test, capable of revealing signs of sleep deprivation. The resulting genetic biomarker blood test can reportedly show with 92 percent accuracy whether a sample is taken from a well-rested or sleep-deprived individual.
“The test uses a machine learning algorithm approach to identify a signature of gene expression levels in a subset of 68 genes from thousands that were measured in a single blood sample,” Simon Archer, a professor of Molecular Biology of Sleep at the University of Surrey, told Digital Trends. “This is exciting because it provides the groundwork toward developing an automated test that could determine if someone was drowsy and unfit to perform critical job-related tasks or be in charge of a vehicle.”
The machine learning tool analyzed samples of blood taken from both rested study participants and those who had skipped one night of sleep. While this is still early stages of the research, it opens up promising avenues of exploration for the future.
“Our results are far from developing a roadside test for drowsiness,” Archer continued. “What we do takes a few days in the lab to measure levels of gene expression in a blood sample. But, in theory, the signature genes that we have identified that predict sleep loss status could form part of an automated test — we just need the technological development to support it.”
A paper describing the work was recently published in the journal Sleep.
