EEG-based stress recognition amongst Universiti Sains Malaysia (USM) students

Sharif, Eizan Azira Mat (2021) EEG-based stress recognition amongst Universiti Sains Malaysia (USM) students. Masters thesis, Univesiti Sains Malaysia.

[img]
Preview
PDF - Submitted Version
Download (579kB) | Preview

Abstract

Introduction: Many students struggle with stress associated with their studies regardless of school, college, or university. Research has revealed that students who have excessive stress will have difficulty focusing on learning, which has a negative impact on academic outcomes that lead to health problems. Purpose: This study aimed to detect stress among undergraduate and postgraduate students from various faculties of Universiti Sains Malaysia (USM), Main Campus, Penang. Methodology: The electroencephalography (EEG) system was used to identify student's brainwave patterns while exposing different stress levels. EEG was chosen because it offers several advantages such as non -invasive data acquisition, ease of use, low-cost preparation, and a high temporal resolution in milliseconds. Besides that, the researcher used the Perceived Stress Scale - the self-assessment instrument, to assess students’ stress levels. In this study, the researcher applied four Stroop Tests to induce stress. Results: The results showed that the alpha and beta waves were the most common higher frequency bands among undergraduate and postgraduate students. The researcher decided to apply the study from Priyanka (2016); therefore, the beta wave was considered the stress detection level. Entropy and Standard Deviation were the accurate classifiers to detect stress levels. Statistical analysis showed the mean values for PSS10 Score undergraduate (n=24) = 21.67 and for postgraduate (n=6) = 21.17 with the p-value of .228. The p-value was greater than 0.05 (p > 0.005), therefore, there were no significant mean differences of the perceived stress scale between undergraduate and postgraduate students from various faculties of Universiti Sains Malaysia (USM), Main Campus (Penang) during stress-inducing tasks. For perceived stress scale score between gender (male and female) revealed that the mean values for male (n=15) = 21.47 and female (n=15) = 21.67 with the p-value of .847, and the pvalue was greater than 0.05 (p > 0.05). As a result, there were no significant differences in perceived stress scores between males and females from various faculties of Universiti Sains Malaysia (USM), Main Campus (Penang) during stress-inducing tasks. The two-way repeated-measures ANOVA for duration revealed no significant difference in the duration of the Stroop tests (F (3, 87)) = 1.860, p =.142, and for between-group interaction showed no significant difference in the duration of the Stroop tests between programs within the four Stroop tests (F (3,84)) = .061, p = .980. Conclusions: It can be concluded that this study that detects the stress level among students using an EEG system could alter the way of detection and treatment of some severe health problems over other current practises. It provided us with a more diverse assessment of stress conditions that might not be possible for one to express. The combination of signal processing techniques such as Wavelet Transform and Coiflet1 with three formulas from Energy, Entropy and Standard Deviation features developed by the time-frequency analysis of EEG signals proved to enhance accuracy.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Electroencephalography
Subjects: R Medicine
Divisions: Kampus Kesihatan (Health Campus) > Pusat Pengajian Sains Perubatan (School of Medical Sciences) > Thesis
Depositing User: Mr Abdul Hadi Mohammad
Date Deposited: 15 Aug 2022 03:43
Last Modified: 15 Aug 2022 03:43
URI: http://eprints.usm.my/id/eprint/54046

Actions (login required)

View Item View Item
Share