Continuous mental stress level assessment using electrocardiogram and electromyogram signals

Stress endangers the mental and physical health of individuals and also has a major impact on society. Accurate, reliable and quantitative measurement of stress leads to detect the person's stress level in earlier stages and reduce the intervention time to manage it. Individuals' stress gradually changes from one level to another. That's why traditional stress recognition approaches are not useful in real-life. Besides, the physiological stress response is subject-dependent. We proposed a continuous personalized stress assessment method. For this purpose, a fuzzy-based model was employed which uses the electrocardiogram and electromyogram signals together to achieve a highly reliable and accurate mental stress index. A total of 34 healthy participants were recruited during a pre-designed stress-inducing protocol. The results of the experiments illustrated the feasibility of the proposed approach. The perceived stress score showed a high correlation (more than 0.9) with the mental stress index. Moreover, the average stress classification accuracy across all subjects for two-level and three-level achieved as 96.7 % and 75.6 %, respectively. The acceptable accuracy of multi-level stress detection, along with the obtained personalized stress index and high correlation between the perceived and estimated stress, emphasize the proposed methodology as an innovative approach, in line with state-of-the-art research.

Automated summary

The study proposed a continuous personalized stress assessment method using a fuzzy-based model combined with electrocardiogram and electromyogram signals to achieve a highly reliable and accurate mental stress index. The experiments demonstrated the feasibility of the approach, with high accuracy in multi-level stress detection and a strong correlation between perceived stress and estimated stress, emphasizing the innovative nature of the methodology in line with current research trends.