Electrocortical Output and Hemispheric Contributions to the Alert State During a Sustained Attention Task

This study investigated the relative electrocortical contributions of specified cerebral locations and the potential motoric influence of handedness in a biocybernetic feedback environment. Thirty six right-handed subjects engaged in two sixteen minute trials of a compensatory tracking task by guiding (via a joystick) a circle over a stationary target presented on a computer screen, performing one trial with their preferred (right) and one using their non-preferred (left) hand. For each subject, a Macintosh computer interpreted electrocortical activity through an engagement index (20*beta/(alpha+theta)) and changed the task mode, i.e. manual (subject guides the task) or automatic (computer guides the task) as a function of electroencephalogram (EEG) and the feedback condition (positive or negative). Continuous EEG and performance, measured as Root Mean Square (RMS) error, were recorded from the subjects during trials. RMS error was used to assess performance differences between preferred/non-preferred hand. Task mode switches were used to analyze differences between positive/negative feedback, preferred/non-preferred hand, and cortical site(s) . These aforementioned variables and differences between manual/automatic task mode were also assessed using the average engagement index. Results indicate that the biocybernetic system is significantly influenced both by the site(s) used to generate the engagement index and the hand used to guide the tracking task. An analysis of RMS error showed that subjects were significantly more proficient in tracking using the preferred hand. Significant differences were found for task mode switches between positive and negative feedback and a significantly higher engagement index was observed for the non-preferred hand condition.