Smartphone sensor data has potential to detect cannabis intoxication

According to a new study by the Rutgers Institute for Health, Health Care Policy and Aging Research, smartphone sensors are very similar to those used in GPS systems to determine if someone is drunk after consuming marijuana. It could be the way.

Studies show Drug and alcohol dependenceAssessed the possibility of using smartphone sensor data to identify episodes of cannabis addiction in the natural environment. The accuracy of the combination of the time function (time and day of the week tracking) and the sensor data of the smartphone was 90%.

“using sensor In a person phoneDetects when a person may be experiencing cannabis addiction and may be able to provide simple intervention at the time and place where it is most likely to have an impact to reduce cannabis-related harm. ” And a professor of psychiatry, the Rutgers Institute for Health, Health Care Policy and Aging Research Center for Population Behavior and Health.

Cannabis addiction is associated with delayed response times and can affect performance at work and school, impair driving behavior, and lead to injury and death. Existing detection methods such as blood, urine, and saliva testing have limitations as indicators of cannabis addiction and cannabis-related disorders. Everyday life..

Researchers analyzed daily data collected from young adults who reported cannabis use at least twice a week. They examine phone surveys, cannabis use self-start reports, and ongoing phone sensor data to determine the importance of time and day of the week in detecting use, and the phones that are most helpful in detecting self-reported cannabis addiction. I have identified the sensor.

They have 60% accuracy in detecting self-reports of cannabis addiction by time zone and day of the week, and 90% accuracy in detecting cannabis addiction by combining time function and smartphone sensor data. I found.

Movement patterns from GPS data (when reportedly feeling high) and movement data from accelerometers that detect various movements were the most important phone sensor functions for detecting self-reported cannabis poisoning. ..

Researchers can use low-burden methods (time and day tracking and analysis of phone sensor data) to detect addiction in daily life and use phone sensors to detect subjective addiction due to cannabis consumption I found that it was highly sexual.

Future research should investigate the performance of the algorithm in classifying drunk and non-drunk reports for people who use cannabis infrequently.Researchers need to study the report drunkenness Using that tool Law enforcement agency May be used to show a stronger correlation with self-reporting Use of cannabis..

Authors of the study include faculty members from Stevens Institute of Technology, Stanford University, Carnegie Mellon University, the University of Tokyo, and Washington University in Seattle.