Technology Category
Cognitive Modeling & Cognitive Systems: Affective ComputingIntelligent Robots: Multimodal Perception & Sensor FusionHumans and AI: Brain-Sensing and Analysis
Regression-Based Approach to Anxiety Estimation of Spider Phobics During Behavioural Avoidance Tasks
🤖 AI Summary
Traditional behavioral avoidance tests (BATs) and self-report questionnaires provide only snapshot assessments of anxiety, failing to capture its dynamic evolution. To address this limitation, we propose a continuous, wearable-based physiological modeling framework for real-time anxiety intensity estimation. Using a wrist-worn device, we synchronously acquire heart rate (HR), heart rate variability (HRV), electrodermal activity (EDA), and skin temperature; extract time-series features via sliding windows; and integrate task-phase contextual information to construct a multimodal physiological-contextual regression model. The method enables unobtrusive, continuous, and quantitative anxiety monitoring throughout the entire BAT procedure in spider-phobic participants. Experimental results demonstrate a root-mean-square error (RMSE) of 0.197 and a mean absolute error (MAE) of 0.041—significantly outperforming baseline approaches. These findings validate the efficacy and clinical potential of context-augmented physiological modeling for dynamic anxiety assessment.
Technology Category
Application Category
📝 Abstract
Phobias significantly impact the quality of life of affected persons. Two methods of assessing anxiety responses are questionnaires and behavioural avoidance tests (BAT). While these can be used in a clinical environment they only record momentary insights into anxiety measures. In this study, we estimate the intensity of anxiety during these BATs, using physiological data collected from unobtrusive, wrist-worn sensors. Twenty-five participants performed four different BATs in a single session, while periodically being asked how anxious they currently are. Using heart rate, heart rate variability, electrodermal activity, and skin temperature, we trained regression models to predict anxiety ratings from three types of input data: (1) using only physiological signals, (2) adding computed features (e.g., min, max, range, variability), and (3) computed features combined with contextual task information. Adding contextual information increased the effectiveness of the model, leading to a root mean squared error (RMSE) of 0.197 and a mean absolute error (MAE) of 0.041. Overall, this study shows, that data obtained from wearables can continuously provide meaningful estimations of anxiety, which can assist in therapy planning and enable more personalised treatment.
Problem
Research questions and friction points this paper is trying to address.
Estimating anxiety intensity during behavioral tasks using physiological data
Improving anxiety prediction by adding contextual task information
Using wearable sensors for continuous anxiety monitoring in therapy
Innovation
Methods, ideas, or system contributions that make the work stand out.
Wrist-worn sensors collect physiological anxiety data
Regression models predict anxiety using multiple features
Contextual task information enhances model accuracy
🔎 Similar Papers
No similar papers found.
