Detecting the Early Drop of Attention using EEG Signal

The capability to detect the drop of attention as early as possible has many practical applications including for the development of the early warning system for those who involves in a high risk work that require a constant level of concentration. This work intends to evaluate the possibilty of detecting the early drop of attention from the brain signals: delta,  theta, alpha, beta, and gamma, waves. The brain waves are obtained from subjects who performs a continunous performance test. In the test, the subjects are asked to hit the space key when an `X’ charackter shows up on their computer screen. If the sreen shows any character but `X’, they should ignore them. The duration between the time the stimulus is displaed and the time the subject hit the space key is used to represent the level of attention. A minimum delay time is associated with a high level of attention. Otherwise, a large delay time is associated with a low level of attention. A k-NN classification method is established with $k = 3$. The results suggest that the best detection of the attention drop can be performed when the attention features are extracted from the earliest stage of the brain wave signals. In addition, the brain wave signal should be recorded longer than 1~s since the stimulus is presented. A significant drop in the level of response time is required prior the brain signal can be used for detecting the attention change.

Fergyanto E Gunawan, Dr Eng and Krisantus Wanandi