Detecting Smoking Events Using Accelerometer Data Collected Via Smartwatch Technology: Validation Study

Abstract

BackgroundSmoking is the leading cause of preventable death in the world today. Ecological research on smoking in context currently relies on self-reported smoking behavior. Emerging smartwatch technology may more objectively measure smoking behavior by automatically detecting smoking sessions using robust machine learning models.ObjectiveThis study aimed to examine the feasibility of detecting smoking behavior using smartwatches. The second aim of this study was to compare the success of observing smoking behavior with smartwatches to that of conventional self-reporting.MethodsA convenience sample of smokers was recruited for this study. Participants (N=10) recorded 12 hours of accelerometer data using a mobile phone and smartwatch. During these 12 hours, they engaged in various daily activities, including smoking, for which they logged the beginning and end of each smoking session. Raw data were classified as either smoking or nonsmoking using a machine learning model for pattern recognition. The accuracy of the model was evaluated by comparing the output with a detailed description of a modeled smoking session.ResultsIn total, 120 hours of data were collected from participants and analyzed. The accuracy of self-reported smoking was approximately 78% (96/123). Our model was successful in detecting 100 of 123 (81%) smoking sessions recorded by participants. After eliminating sessions from the participants that did not adhere to study protocols, the true positive detection rate of the smartwatch based-detection increased to more than 90%. During the 120 hours of combined observation time, only 22 false positive smoking sessions were detected resulting in a 2.8% false positive rate.ConclusionsSmartwatch technology can provide an accurate, nonintrusive means of monitoring smoking behavior in natural contexts. The use of machine learning algorithms for passively detecting smoking sessions may enrich ecological momentary assessment protocols and cessation intervention studies that often rely on self-reported behaviors and may not allow for targeted data collection and communications around smoking events.

State Transition Modeling of the Smoking Behavior Using LSTM Recurrent Neural Networks

“…It also allows for "in-time" intervention techniques to be developed. Our previous work [6] in this field has indicated that smoking can be detected using only accelerometer data and single-layer artificial neural networks (ANNs) with an accuracy of ~95% in laboratory settings and ~90% in real-world settings [6], [11]. Independent reports [12]- [16] also confirmed the usability of smartwatches in the study of human behavior.…”

“…Smartwatches have been used as stepcounters [16], sleep monitoring [18], diet monitoring [13] as well as general fitness tracking [19]. In the context of smoking, smartwatches have demonstrated to be usable for in-situ study of smoking [20], [21] with high accuracy [6], [20]- [22]. Smartwatches have been used to detect smoking gesture with 95% accuracy in laboratory environment [23] and 90% in-situ detection of smoking [6].…”

“…In the context of smoking, smartwatches have demonstrated to be usable for in-situ study of smoking [20], [21] with high accuracy [6], [20]- [22]. Smartwatches have been used to detect smoking gesture with 95% accuracy in laboratory environment [23] and 90% in-situ detection of smoking [6]. Study of smoking has also been demonstrated to be more accurate when compared to selfreporting (90% versus 78%) [6], [24].…”

“…Smartwatches have been used to detect smoking gesture with 95% accuracy in laboratory environment [23] and 90% in-situ detection of smoking [6]. Study of smoking has also been demonstrated to be more accurate when compared to selfreporting (90% versus 78%) [6], [24].…”

“…Most studies conducted rely on participants to self-report their smoking behavior. In various studies [2], [6], [7], the accuracy of self-reporting has been shown to be no more than ~76%. To bypass the reliance on self-reporting, some studies have been conducted in laboratory-based settings.…”