Machine Learning Prediction of Smoking Risk from Repeated Sampling of Environmental Images (Preprint)

BACKGROUND Viewing their habitual smoking environments increases smokers’ craving and smoking behaviors in laboratory settings. A deep learning approach can differentiate between habitual smoking versus nonsmoking environments, but its ability to predict smoking risk associated with the broader range of environments smokers encounter in their daily lives is unknown. OBJECTIVE To predict environment-associated smoking risk from continuously acquired images of smokers’ daily environments. METHODS Smokers from the Durham, NC area completed ecological momentary assessments both immediately smoking and at randomly selected times throughout the day, for two weeks. At each assessment, participants took a picture of their current environment and completed a questionnaire on smoking, craving, and the environmental setting. A convolutional neural network (CNN)-based model was trained to predict smoking, craving, whether smoking was allowed, and whether the participant was outside based on images of participants’ daily environments, the time since their last cigarette, and baseline data on daily smoking habits. Prediction performance, quantified using the area under the receiver operating characteristic curve (AUC) and average precision (AP), was assessed for (a) out-of-sample prediction, and (b) personalized models trained on images from days 1-10. Models were optimized for mobile devices and implemented as a smartphone app. RESULTS Forty-eight participants completed the study, and 8008 images were acquired. The personalized models were highly effective in predicting smoking risk (AUC=0.827; AP=0.882), craving (AUC=0.837; AP=0.798), whether smoking was allowed in the current environment (AUC=0.932, AP=0.981), and whether the participant was outside (AUC=0.977, AP=0.956). The out-of-sample models were also effective in predicting smoking risk (AUC=0.723, AP=0.785), whether smoking was allowed in the current environment (AUC=0.815, AP=0.937), and whether the participant was outside (AUC=0.949, AP=0.922); but were not effective in predicting craving (AUC=0.522, AP=0.427). Omitting image features reduced performance (p<0.05) when predicting all outcomes except craving (p>0.05). Smoking prediction was more effective in participants whose self-reported location type was more variable (Spearman’s ρ=0.48, p=0.001). CONCLUSIONS Images of daily environments can be used to effectively predict smoking risk. Model personalization, achieved by incorporating information about daily smoking habits and training on participant-specific images, further improves prediction performance. Environment-associated smoking risk can be assessed in real time on a mobile device, and could be incorporated in device-based smoking cessation interventions.

P5301Impacts of lifestyle behavior and shift Work on visceral fat accumulation and progression of atherosclerosis in middle-aged workers

Background Work style, and particularly shift work, can affect an individual's health through disrupting circadian rhythms. Moreover, lifestyle habits including dietary and exercise routines might be altered by irregular shift hours. We thus hypothesized that an individual's lifestyle including working habits could influence the prevalence of visceral fat obesity and the progression of atherosclerosis. Purpose The present study investigated how lifestyle and shift work affect the accumulation of visceral fat and the progression of subclinical atherosclerosis in middle-aged workers. Methods This study enrolled employees undergoing their periodic health check-up (n=10883). The Cardio-Ankle Vascular Index (CAVI) was measured to assess arterial stiffness, followed by ultrasound examination and computed tomography imaging to measure carotid intima-media thickness (IMT) and visceral fat area (VFA), respectively. Lifestyle was evaluated by the following items: 1) eating breakfast, 2) nighttime eating, 3) regular exercise, 4) habitual drinking, 5) habitual smoking, 6) sleeping hours, and 7) working hours. With regard to work factors, subjects were categorized into fixed daytime workers or shift workers (including subjects working with an irregular schedule, outside of daytime hours, or at nighttime). Results Among all subjects enrolled, 6820 subjects were fixed daytime workers and 4063 subjects were shift workers. Most of the employees engaged in fixed daytime work were clerical workers, while the employees engaged in shift work were mainly physical workers in this company. Fixed daytime workers had significantly greater VFA than shift workers, but the prevalence of metabolic syndrome, CAVI values, and carotid IMT were similar between groups. Reduced regular exercise, long sleeping hours, and fixed daytime work were independently associated with visceral fat accumulation by both multivariate regression and logistic regression analyses. However, the logistic regression analysis with the presence of metabolic syndrome as the endpoint revealed that skipping breakfast, reduced regular exercise, long sleeping hours, and short working hours were independent determinants of metabolic syndrome. On the other hand, univariate and multivariate regression analysis showed that habitual smoking, but not shift work, were significantly associated with CAVI and carotid IMT. Logistic regression analysis with the endpoint of carotid atherosclerosis (presence of plaque) showed that habitual smoking was an independent determinant of carotid atherosclerosis. Conclusions Reduced regular exercise, long sleeping hours, and fixed daytime work were significantly associated with visceral fat accumulation, while habitual smoking has a consistent association with the progression of atherosclerosis. These findings support the concept that unhealthy lifestyles should be modified before considering intervention in work styles.