Leveraging Mobile Phone Sensors, Machine Learning, and Explainable Artificial Intelligence to Predict Imminent Same-Day Binge-drinking Events to Support Just-in-time Adaptive Interventions: Algorithm Development and Validation Study

“Smartphone sensors provide potentially powerful tools for collecting continuous data that can be used to infer personal behaviors or activities associated with alcohol use with low burden and at low cost.” p.2

Bae et al., 2023

I found this article to be particularly interesting, as artificial intelligence is on the cusp of being utilized for many facets. This study takes place in an urban area of western Pennsylvania, with 75 individuals between ages 21-25, who have had at least one binge drinking event (BDE) reported by emergency departments and screened positive for hazardous drinking. It was fascinating to learn that the combination of artificial intelligence, machine learning, and the use of phone sensors could be leveraged to predict and support interventions with binge drinking at a 95% rate on weekends and a 94% rate on weekdays. The low burden and low cost associated with utilizing mobile phones could be a new way to address this concern, and while participants gave informed consent to data collection, I worry about privacy concerns with a prediction model such as this. How is this data being protected? Who will be able to access it? How do we know participants are being honest when they are self-reporting consumption habits? Despite privacy concerns, I would be curious to further explore the connection between these smartphone sensors and behaviour.

Researchers were able to draw various conclusions from this study: “In predicting drinking events, we found that young adults who traveled more (eg, larger radius of gyration) and spent longer duration in important locations were more likely to report BDEs (ie, compared with non-drinking events) and drinking events (ie, non-drinking and drinking events) later that day. Furthermore, the participants who interacted with their smartphones more (ie, higher acceleration values) and had longer call durations were less likely to drink alcohol later in the day. These results suggest that smartphone interaction activity and communication during the day might indicate work, school, and social obligations and activities that could help to regulate or constrain drinking behavior. By contrast, young adults who had fewer communication interactions using the phone had an increased likelihood of drinking later that day. The low level of communication activity involving the phone in the hours before a drinking event may reflect a sense of social isolation, which is a direction to be examined in future research. The most important smartphone sensor features contributing to drinking behavior prediction may serve as early warning signals, which have parallels in the mental health literature, where, for example, certain signs and symptoms signal increases in depressive symptom severity.” p.15

Bae SW, Suffoletto B, Zhang T, Chung T, Ozolcer M, Islam MR, Dey AK. Leveraging Mobile Phone Sensors, Machine Learning, and Explainable Artificial Intelligence to Predict Imminent Same-Day Binge-drinking Events to Support Just-in-time Adaptive Interventions: Algorithm Development and Validation Study. JMIR Form Res. 2023 May 4;7:e39862. doi: 10.2196/39862. PMID: 36809294; PMCID: PMC10196900.


( Average Rating: 5 )

One response to “Leveraging Mobile Phone Sensors, Machine Learning, and Explainable Artificial Intelligence to Predict Imminent Same-Day Binge-drinking Events to Support Just-in-time Adaptive Interventions: Algorithm Development and Validation Study”

  1. kgear

    Predictive technology to increase socioemotional and physiological health based on real-world, real-time events is both enticing and questionable. In Bae’s specific 2023 research, I wonder whether the correlation made between high mobile phone technology communication by day and low binge drinking events by night, and vice versa, is verifiable causation or if there are other factors missing in the analysis. As compelling as it sounds for avid phone users, the research team seems to activate an excess of leaps in logic to arrive at a specific end game scenario of blaming low phone usage for binge drinking. Should the research prove to be credible, which, then, is worse: alcohol or algorithm addiction? Drinks or clicks? While I am certainly not advocating the use or abuse of alcohol, the social isolation resulting from mobile phone usage may be more detrimental to one’s final aim in life (as varied as may be) than Firewater. That, of course, would also depend on a variety of individual characteristics of one’s disposition. These personal factors seem like they ought to have more bearing on alcohol intake than phone use. In fact, a quick Google search reveals myriad of ways in which mobile phones can serve alcohol reduction.
    Dr. Vogt reminded us in Week 1: “The Internet was first about information (a giant brochure rack), then about interaction (click to watch something happen), and is now about participation (the “social web” where value is generated primarily by users). It is addictive because participation breeds identity. But is it really “social” when no other real people are tangibly nearby?” I am absolutely in awe, and also kinda terrified, of where this research will go.


    ( 0 upvotes and 0 downvotes )

Leave a Reply

You must be logged in to post a comment.