Authors:

Alpamys Sultanbek

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Stress is on an upward trend among young adults-with the pressure to succeed increasingly affecting students (1). This growth in stress has been a major contributing factor in numerous health issues and could be linked to the growing rate of suicide among youth (2) (3). A major component and cause of stress are timed, graded and quantified exams, especially because these exams hold great importance in achieving career success. This principle has been shown in a previous 2022 study-successfully inducing stress in participants using a timed math quiz (4). Based on these findings, the present experiment investigated a method of stress diagnosis using brain activity. The 2022 study's methodology was used to induce stress in 12 college students. Various biometrics, including: skin temperature, heart rate, and brain signal, were collected while participants were performing a relaxing activity or a timed quiz. This data was then used to train an ML model to recognise stress markers in individuals. A Convolution Neural Network trained with only EEG Signal data, given a 5- second snippet of brain activity, was able to predict participant stress with an accuracy of 95%. This method of identifying stress through brain activity can be used alongside other biofeedback methods to improve the diagnostic aspects of stress therapy. For example, in a guided meditation session aiming to reduce stress, brain activity can be measured to determine when a subject has achieved a state of low stress. Keywords: EEG; Biofeedback; Machine Learning; Stress; Neurotech

Source:

Purdue University / 2024

Topics:

No topics listed

Co-authors:

Alpamys Sultanbek