Single trial variability of event-related brain potentials as an index of neural efficiency during working memory

Abstract

Event-related brain potentials (ERPs) to a particular stimulus are extracted from the continuous electroencephalogram (EEG) through signal averaging techniques. The most extensively studied ERP component, P300 (P3, or P3b), occurs at approximately 300–800 ms post-stimulus. P3 amplitude and latency are markers of the attentional/cognitive resources devoted to the task and the timing (within msecs) of central processing speed of an individual’s cognitive response to a stimulus, respectively. Variability in the timing and amplitude of components in the single EEG trials that contribute to the averaged ERP has been of particular interest to our laboratory because it may provide an index of central information processing efficiency. Examination of single trial variability (STV) can provide a level of analysis beyond traditional ERP measures and offers a unique marker of the functional integrity of neural pathways. In the present study we examined ERP STV as it relates to WM demand or load during a visual nback task in normal adult participants. Performance measures and the scalprecorded EEG were obtained during the n-back task. Frontal and parietal scalp averaged ERPs were derived from the EEG time-locked to the stimuli. The mean and variability (SD) of the peak amplitude and latency were then obtained from the single trial data for each participant and condition. Results showed that as WM load increased, behavioral measures of processing speed slowed, behavioral efficiency decreased, and the number of correct responses decreased. Correlations for both latency and amplitude between the P3 component derived from the averaged ERP and the P3 component derived from the single trials were generally high, indicating that the P3s for individual trials identified by our STV procedure were representative of the P3s in the averaged ERP obtained by standard signal averaging procedures. P3 STV analyses also showed differential effects between frontal and parietal scalp sites for both amplitude and latency variability that were related to WM load. Both frontal P3 latency STV and amplitude increased as WM load increased, indicating decreased neural efficiency associated with an increase in WM load. Single trial ERP variability measures may provide potential physiological markers of the neural efficiency of brain processes engaged in cognitive functions, such as working memory.