Detecting Early-Stage Stress Levels Using EEG and Heart Rate Monitoring with Enhanced Sparse Fire Hawk Graph Probabilistic Spiking Attention Neural Network

Electroencephalogram (EEG) is a non-invasive technique that records brain electrical activity, aiding in diagnosing neurological disorders like epilepsy and sleep disorders, while heart rate monitoring measures heartbeats per minute to provide insights into cardiovascular health. Chronic stress, a significant health concern, contributes to various physical and mental health issues, and current stress management solutions often lack personalized interventions due to insufficient real-time physiological data. To address these challenges, this study proposes an advanced system that integrates EEG and heart rate data using an Enhanced Sparse Fire Hawk Graph Probabilistic Spiking Attention Neural Network (ESFHG-Probspi-AtNet) for early-stage stress level detection. Data from 34 individuals were collected and pre-processed using Self-Guided Filtering (SGF) to remove noise, followed by feature extraction using the Gated Position-Sensitive Axial Attention Transformer (GPAAT). Finally, these data are classified using Enhanced Sparse Fire Hawk Graph Probabilistic Spiking Attention Neural Networks (ESFHG-Probspi-AtNet) for detecting stress levels in various stage. The ESFHG-Probspi-AtNet model was implemented on the Python platform and evaluated using four datasets, achieving remarkable results with 99.8% recall and 99.9% accuracy, outperforming existing methods. These results demonstrate the model's exceptional effectiveness and room for advancement in customized stress-reduction strategies.