Mental Stress Detection Using Physiological Sensors and Artificial Intelligence: A Review

Area of Science:

• Biomedical Engineering
• Computational Neuroscience
• Psychophysiology

Background:

• Persistent stress can lead to significant mental and physical health disorders.
• Effective early intervention requires accurate and timely monitoring of mental stress levels.
• The detection of mental stress is a rapidly evolving field with significant clinical implications.

Purpose of the Study:

• To investigate the primary goals and mechanisms of existing mental stress detection frameworks.
• To examine the role of physiological sensors, stressors, algorithms, monitoring methods, and validation tools in stress assessment.
• To highlight advancements in AI-driven approaches for stress detection.

Main Methods:

• Review of various physiological sensors, including wearable options like GSR, ECG, PPG, EEG, and pupil diameter cameras.
• Examination of AI-based methods, particularly machine learning and ensemble models, for interpreting complex physiological data.
• Analysis of stressors used for inducing physiological responses and validated self-report instruments (e.g., STAI, PSS-10) for benchmarking.

Main Results:

• Wearable sensors offer continuous monitoring of physiological responses in real-world environments, revealing stress patterns.
• AI and machine learning models enhance the accuracy of stress classification by interpreting diverse physiological data.
• Self-report instruments correlate strongly with physiological outcomes, serving as benchmarks for sensor-based measurements.

Conclusions:

• Future research should focus on advanced AI and sophisticated sensors for more precise stress level definition.
• Further investigation is needed into understudied physiological factors contributing to stress.
• Integrating multi-sensor data with AI holds promise for improved mental stress detection and management.