Loading...
Wearable Real-time Mental Stress Detector
Abdul Kader, Lamis
Abdul Kader, Lamis
Date
2022-11
Authors
Type
Thesis
Degree
Description
A Master of Science thesis in Biomedical Engineering by Lamis Abdul Kader entitled, “Wearable Real-time Mental Stress Detector”, submitted in November 2022. Thesis advisor is Dr. Hasan Al-Nashash and thesis co-advisors are Dr. Usman Tariq and Dr. Fares Al-Shargie. Soft copy is available (Thesis, Completion Certificate, Approval Signatures, and AUS Archives Consent Form).
Abstract
Mental stress causes physical diseases in addition to behavioral, emotional, and monetary complications. Thus, research is set on detecting mental stress to reduce the risk of damage to an individual’s well-being. There is plenty of evidence and solutions in the literature that help assess and detect stress, highlighting its importance. Assessment and detection of stress can be performed using many physiological signals, including the Electroencephalogram (EEG) and the Galvanic Skin Response (GSR). Moreover, many commercialized systems used to detect stress with EEG require a controlled environment with many channels, which prohibits its daily use. Those systems are also complex and expensive. Fortunately, there is a rise to using wearable devices for monitoring stress, which offers more flexibility to monitor stress through physiological signals. In this thesis, a wearable-monitoring system that integrates both EEG and GSR physiological signals was developed. The novelty of the proposed device is that it requires only one channel for acquiring both EEG and GSR signals. By sensor fusion, we were able to achieve improved accuracy, lower cost, and easier to use device. Power spectrum analysis and machine learning were applied on the acquired signals to detect the elevation of and classify mental stress. Furthermore, the optimum electrode location on the scalp was investigated for stress detection using one channel. This was achieved by utilizing a specially designed mechanical framework with a rail-like structure to allow the flexibility of electrode positioning. The proposed system was tested on 20 human subjects. Results demonstrate the capability of the system in classifying 2 levels of mental stress with a maximum accuracy of 70.3% when using EEG, 93% when using GSR, and 84.6% when using both EEG and GSR.