Insights into Physical Stress: EOG-Based Detection and Classification

EOG measures electrical activity of the eye, which is beneficial in identification of several physiological states and issues concerning the eyes. This work includes the variability of stress on the human body and how it affects ocular activity. Data is first taken for a normal state for ten seconds and then physical stress is applied by splashing some ice water on the face. After feature extraction from the EOG data, machine learning classifiers were used for stress classification into normal, low, medium, and high. The random forest classifier achieved the accuracy of 69.23%. In future, for monitoring and detecting stress in real-time, this work can be incorporated into various diagnostic and therapeutic systems used to manage stress.