Smartwatch-Based Unobtrusive Continuous Anxiety Tracker for Evaluating Post-Stroke Patients’ Quality of Life

Abstract

Post-stroke anxiety (PSA) affects 20–30% of stroke survivors and significantly impacts quality of life (QoL) and rehabilitation outcomes. Traditional emotion assessment relies on subjective self-reports, limiting to capture real-time fluctuations in emotional states. This study proposes a self-supervised learning (SSL) framework combined with a transformer-based emotion classifier to enable continuous anxiety tracking in patients with stroke using smartwatch-derived photoplethysmography (PPG). The SSL model was pretrained on the VitalDB dataset using R-peak-to-peak intervals (RRIs) extracted from electrocardiography (ECG) signals. The pretrained encoder is then integrated into a transformer-based classifier trained on the Psychophysiology of Positive and Negative Emotions (POPANE) dataset with labeled emotional responses. The trained model was applied to smartwatch-derived pulse peak intervals (PPI) from patients with stroke, and evaluated against Generalized Anxiety Disorder-7 (GAD-7) scores and we evaluated it against GAD-7 scores using both group-level and 90-day longitudinal analyses, along with on-device feasibility on a Galaxy Watch6. Over the 30 days preceding the second survey, between-group differences were significant by Welch’s t-test (p = 0.0086), and discrimination reached an AUC of 0.859 with a 95% confidence interval of 0.664–0.992. In 90-day monitoring, generalized estimating equations showed persistent divergence when groups were defined by the second survey, with significant differences across multiple weeks preceding the survey, consistent with the retrospective GAD-7 window. These findings indicate that ECG-pretrained cardiac representations can be translated to smartwatch PPG for unobtrusive, continuous anxiety tracking in stroke. The results provide feasibility-level evidence and motivate larger, multi-site prospective validation toward clinical deployment.