Exploring Sensing Technologies for Tracking Healthy Eating Behaviors: Systematic Review

This is the authors' abstract. We don't add key points for this paper.

Sensor-based technologies for healthy eating have gained attention in human-computer interaction research. With the prevalence of diet-related health issues such as obesity and diabetes, innovative interventions providing personalized dietary feedback are increasingly needed. Sensor-based technologies offer real-time monitoring of dietary habits. We reviewed 28 articles, published between 2014 and 2024, to uncover the state-of-the-art and research gaps. Sixteen articles (57.1%) reported on effectiveness: seven measured by machine learning model performance and nine through user study evaluation. We identified several eating behaviors that were supported: adequate food intake, meal size and quality, eating posture, swallowing sounds and difficulties, eating styles, and pace. These behaviors were associated with conditions such as diabetes, post-stroke issues, cardiovascular diseases, obesity, and binge eating. Limitations included little to no interventions for life-threatening conditions. Our findings present opportunities to develop tailored sensor-based interventions for various diet-related issues.