An artificial intelligence model can detect if someone has recently used cannabis, according to scientists who say this could help quickly assess if someone is dangerously intoxicated and may need medical care.
Experts concur that urine, saliva and hair strand tests can ascertain if someone has used cannabis, but can take a long time to analyze because marijuana does not immediately show up in some bodily fluids.
What Did Scientists Do To Remedy This Delay? AI To The Rescue.
Sang Won Bae at Stevens Institute of Technology in New Jersey and her colleagues came up with a solution, or so it seems.
They studied 33 adults who used cannabis at least twice a week and reported their use every day for up to 30 days.
With AI tracking and interpreting data from the activity tracker and sensors, medical professionals say they can look at the stored data and detect when any unusual activity started, which could suggest when someone became intoxicated. “It’s about preventing harm from happening and keeping people safe,” said Tammy Chung, one of the Rutgers University researchers.
Are We There Yet? Not Quite.
Joseph Wu at Stanford University and Mark Chandy at Western University in Canada concur that AI requires testing in a larger group of people. Researchers also did not report how much tetrahydrocannabinol the participants consumed or how they consumed it (smoke or oral ingestion), said Chandy, noting that each method can affect a person’s degree of cannabis intoxication.
Chung, according to the outlet, said the accuracy of the AI depended on how the participants self-reported their cannabis use. “If they’re not reliably reporting on that, that could be a problem for the [AI] model.” She said the drug effects questionnaire, which asks people to report how they feel after taking a substance, could be a more quantitative way of assessing cannabis intoxication.
During that period, volunteers wore an activity tracker to collect information on their heart rates, step counts and sleep quality. Sensors on the participants’ phones also provided information on their micromovements, such as how they held their phone and moved their arms, to gauge their stability and coordination, wrote New Science.
What Did They Find?
Data from some of these participants was used to train AI to detect if someone consumed cannabis, then researchers tested the trained AI on the participants’ data. Taking into account false positives and negatives flagged by AI, it was 85% accurate at detecting someone who’d gotten moderately stoned on cannabis within the last 5 minutes.