Smartphone sensor data may detect cannabis poisoning

According to this research, published in ``Drug and Alcohol Dependence, while evaluating the feasibility of using smartphone sensor data to determine that cannabis is intoxicated in the natural environment, the combination of time characteristics (tracking the time of day and the week of the day) and the smartphone The sensor data has 90% accuracy

“By using a person’s cell phone sensor, we may be able to detect when a person may be experiencing cannabis poisoning and conduct brief interventions at the time and place where it may have the greatest impact on reducing cannabis-related harm,” said corresponding author Tammy Chung

Cannabis poisoning is related to slow response, affecting work or school performance, or impairing driving behavior, resulting in injury or death

The researchers analyzed daily data collected from young people who reported smoking marijuana at least twice a week

They found that the time of day and day of the week are 60% accurate in detecting cannabis poisoning self-reports, and the combination of time characteristics and smartphone sensor data is 90% accurate in detecting cannabis poisoning

Travel patterns from GPS data—sometimes when they report feeling excited—and movement data from accelerometers that detect different movements are the most important cell phone sensor characteristics for detecting self-reported cannabis poisoning

Researchers used low-burden methods (tracking daily and weekly time and analyzing cell phone sensor data) to detect poisoning in daily life, and found that using cell phone sensors to detect subjective poisoning of marijuana consumption is very feasible

Future research should investigate the performance of the algorithm in classifying drunken and non-drunken reports of people who do not use marijuana regularly

The authors of the study included faculty from Stevens Institute of Technology, Stanford University, Carnegie Mellon University, the University of Tokyo in Japan, and the University of Washington in Seattle

Journal Reference :

1. Sang Won Bae, Tammy Chung, Rahul Islam, Brian Suffoletto, Jiameng Du, Serim Jang, Yuuki Nishiyama, Raghu Mulukutla, Anind Dey.