Analysis of EEG signal to investigate the influence of intoxication in perception delay

Abstract

Human brains receive a variety of raw data through sense organs, and process them to generate meaningful information like an image or an alarm etc. This mapping depends on personal experiences and training. Environmental information is continuously passed to and integrated in cerebral cortex, where it is evaluated on the basis knowledge acquired from previous experiences. Most of the time, the environment around us is mundane. Occasionally, the situation is alarming warranting immediate attention or a special action to achieve a reward. We are aware of an imminent danger or possible award. Under such a situation, a strong electrical signal is generated. However, it takes time to integrate external stimuli with pre-acquired knowledge to perceive that an action is needed. So, there is a time lag between external stimuli to its perception (Tper) and perception to action (Tact)· This time lag varies with age and physical condition. For example, an old person is slower than a young one, and someone intoxicated or drowsy will be slower than one physically fit and fresh. The purpose of this work is to define the Perception Delay such that it is reliable and stable. We took frequency features of EEG signals on slotted time over several probes, and integrate them to a feature vector. By unsupervised clustering, we could find the exact time when the signal behavior changes abruptly, indicating onslaught of the powerful signal. Through experiments we ensured confidence of our definition. In addition, we also measured perception delay after getting the subjects intoxicated and compared it when the subjects are normal.