Genetic Algorithm-Based Human Mental Stress Detection and Alerting in Internet of Things.

Area of Science:

• Healthcare technology
• Psychophysiology
• Machine learning in medicine

Background:

• Chronic stress negatively impacts emotional well-being and can lead to health issues.
• Early stress detection through continuous monitoring is crucial for prevention.
• Wearable devices offer real-time data collection for stress assessment.

Purpose of the Study:

• To investigate stress detection using physiological signals and deep learning.
• To develop a system for early identification of mental stress.
• To alert patients and doctors about critical stress levels.

Main Methods:

• Utilized electroencephalography (EEG), photoplethysmography (PPG), and galvanic skin response (GSR) sensors.
• Employed a genetic algorithm for feature extraction and ECNN-LSTM for classification.
• Applied preprocessing techniques to remove signal artifacts and used the DEAP dataset.

Main Results:

• The developed system effectively detects stress beyond a threshold, triggering an alert.
• Performance was evaluated against traditional methods using accuracy, F1-score, precision, and recall.
• The system successfully classifies stress levels in various conditions like traveling and learning.

Conclusions:

• The proposed deep learning approach with physiological sensors shows promise for accurate stress detection.
• Real-time stress monitoring and early alerts can aid in managing mental well-being.
• This technology has significant potential for integration into Internet of Things (IoT) healthcare applications.