Mental stress classification during a motor task in older adults using an artificial neural network

UbiComp/ISWC '20: 2020 ACM International Joint Conference on Pervasive and Ubiquitous Computing and 2020 ACM International Symposium on Wearable Computers Virtual Event Mexico September, 2020(2020)

引用 7|浏览7
All people cope with mental stress from time to time. Stress can affect our emotional and physical health, which can lead to physical and/or mental health issues. Our experiment aimed to derive the stress levels of 57 older adults from the electrocardiogram (ECG) signal during a lab study that involved a hang-grip strength task. This experiment bridges the gap between previous studies by classifying the mental stress state of older adults while performing a motor task before and after the stressor was induced. In this study heart rate and heart rate variability multi-dimensional features in the time-, and frequency-domain are extracted and an optimized Artificial Neural Network (ANN) created to identify two states --- stress, or no-stress. We achieved accuracy of 90.83%.
Machine learning, Elderly population, Stress detection, ECG signal, Artificial Neural Network
AI 理解论文
Chat Paper