What is the clinical evidence from this study?

Study design: Other. Population: Stress (n=11). Intervention: Machine learning classification using EEG and peripheral physiological signals vs. Individual physiological signals alone. Primary outcome: Accuracy in identifying stress type and stress level.

October 24, 2024Open Access

Identifying neurophysiological correlates of stress

Key Result

A machine learning approach fusing features from EEG and peripheral physiological signals achieved an average accuracy of 98.1% and 97.8% in identifying stress type and stress level, respectively, using 4-second signal durations.

Structured PICO

Does fusing features from multiple physiological signals including EEG improve the accuracy of multi-type and multi-level stress detection in healthy individuals?

Population

11 healthy young adults (mean age 24.7 years, 27.3% female) participated in validated stress-inducing protocols to evaluate multi-level and multi-type stress detection using physiological signals.

Intervention

Machine learning-based multi-type and multi-level stress detection using fused features from electroencephalography (EEG) and peripheral physiological signals (HR, SpO2, temperature, GSR, PPG) analyzed over ultra-short durations (4s to 60s)

Comparator

Detection using peripheral physiological signals alone or EEG signals alone; non-stress conditions

Outcome

Classification accuracy for identifying stress type (social vs. mental vs. non-stress) and stress levelsurrogate

Fusing EEG and peripheral physiological signals enables highly accurate real-time detection of stress types and levels using ultra-short (4-second) data segments.

Limitations

Small sample size of 11 subjects
Individual physiological variations may affect EEG precision

Abstract

Stress has been recognized as a pivotal indicator which can lead to severe mental disorders. Persistent exposure to stress will increase the risk for various physical and mental health problems. Early and reliable detection of stress-related status is critical for promoting wellbeing and developing effective interventions. This study attempted multi-type and multi-level stress detection by fusing features extracted from multiple physiological signals including electroencephalography (EEG) and peripheral physiological signals. Eleven healthy individuals participated in validated stress-inducing protocols designed to induce social and mental stress and discriminant multi-level and multi-type stress. A range of machine learning methods were applied and evaluated on physiological signals of various durations. An average accuracy of 98.1% and 97.8% was achieved in identifying stress type and stress level respectively, using 4-s neurophysiological signals. These findings have promising implications for enhancing the precision and practicality of real-time stress monitoring applications.

Bookmark

View Full Paper