Detecting Ethanol Intoxication and Impairment Using Wearable Biosensors.

Area of Science:

• Biomedical Engineering
• Forensic Science
• Physiology

Background:

• Objective, real-time measures of alcohol intoxication are lacking for public use.
• Wearable biosensors offer a potential solution for ubiquitous, on-demand intoxication assessment.

Purpose of the Study:

• To evaluate the feasibility of using wrist-worn biosensors to assess ethanol intoxication.
• To determine if physiological data from biosensors can accurately indicate alcohol impairment.

Main Methods:

• Twenty-eight healthy participants were monitored using wrist-worn biosensors measuring heart rate, skin temperature, electrodermal activity, and accelerometry.
• Participants consumed alcohol in a controlled laboratory setting, with regular breath alcohol content and field sobriety tests conducted.
• Machine learning (XGBoost model) was employed to analyze biosensor data for impairment detection.

Main Results:

• Statistically significant changes in physiological parameters (heart rate, skin temperature, electrodermal activity, accelerometry) were observed between sober and intoxicated states.
• The developed XGBoost model achieved a high accuracy of 0.80 in identifying alcohol impairment.
• Continuous physiological monitoring via biosensors proved effective in distinguishing intoxication levels.

Conclusions:

• Wrist-worn biosensors are a feasible tool for assessing ethanol intoxication.
• Physiological data combined with machine learning can provide objective measures of alcohol impairment.
• This technology holds promise for real-world, on-demand intoxication assessment.