WeedGait: Unobtrusive Smartphone Sensing of Marijuana-Induced Gait Impairment by Fusing Gait Cycle Segmentation and Neural Networks

Ruojun Li1, Emmanuel Agu, Ganesh Balakrishnan1, Debra Herman, Ana Abrantes2, Michael Stein2, Jane Metrik

  • 1Worcester Polytechnic Institute
  • 2Butler Hospital

Details

12:30 - 14:30 | Thu 21 Nov | Upper Foyer Balcony | B1P-B.2

Session: Poster Session - Health and Wellness Across the Lifespan 2

Full Text

Digital Library Website

Abstract

Marijuana impairs the motor skills of users, making Driving Under the Influence of Marijuana (DUIM) a growing public health concern. In this paper, we investigate whether dose-dependent changes in participants gait (walk) can be detected using passively gathered smartphone accelerometer and gyroscope data. We envision WeedGait, a smartphone sensing system that will passively detect marijuana users too impaired to drive safely. Our gait analysis utilized normalized, gait segments and could discriminate subjects after smoking either marijuana with 3% or 7.2% THC versus smoking a placebo marijuana cigarette with an accuracy of 92.1%.