Dynamic Neurological Insights: Real-Time Stress Analysis Utilizing Portable EEG Kits

Abstract Clinical Electroencephalogram(EEG) devices are effective in hospitals, but for everyday monitoring, portable options are more accessible and convenient, albeit less accurate. In this research, the working of three portable devices -Neurosky MindWave Mobile 2, Muse 2 headband and Emotiv Insight that have different channels 1,4,5 respectively are explored. They are used to understand students’ state of stress in two different scenarios-listening to online lecture videos (Scenario 1) and live lecture classes in their classrooms (Scenario 2). Twelve Engineering college students participated in the experiment and were asked to watch 10 video lectures from Coursera each video is 4 to 8-minute duration and their EEG data was collected with all the three devices. The same participants were engaged in EEG data collection during their live lecture classes for one hour. A questionnaire consisting of a Likert scale like Self-Assessment Manikin (SAM) was prepared to assess the levels of arousal and valence experienced by the participants. The collected EEG signals from both scenarios are tested with the pre-trained model (Artificial neural network (ANN )model trained using DEAP (A dataset for emotion analysis using EEG, physiological and video signals dataset) and the results of real-time data are analysed. In Scenario 1 which is short duration, EEG recordings showed an accuracy of 55% in Neurosky and 63% in Muse and 76% in Emotiv Insight and in Scenario 2 which is long duration EEG recordings showed an accuracy of 47% in Neurosky and 60% in the muse and 67% in Emotiv Insight. Emotiv shows high accuracy, but Muse proves more comfortable for extended use due to its consistent accuracy in both scenarios.