Bruce McCandliss, Attentional networks

Posner and Petersen (1990) proposed that attention could be understood as three systems supported by separable brain networks responsible for alerting, orienting, and executive skills associated with conflict. The Attention Network Task (ANT) was designed to assess these subcomponents of attention by examining reaction times under several contrasting conditions within the same integrated task.

A simple RT is collected in responseto a left or right central arrow that appears in the periphery. Two flanker conditions are contrasted to tap the conflict network (RT incongruent- RT congruent flankers), while a set of cue conditions are designed to tap the alerting network (RT baseline - RT warning cue) and the orienting network (RT warning cue - RT valid spatial cue). This task has been demonstrated to be a reliable index of these three attentional processes, and these manipulations have been shown to impact brain activity in separable regions associated with conflict, alerting, and spatial orienting.

The current talk reviews recent work extending the ANT to examine mild TBI.Consistent with the majority of theliterature, reaction times for mTBI subjects are significantly slowed across all conditions, and mTBI subjects demonstrate increased variability both within individual subject's responses within a session and across subjects, which introduces significant challenges for analysis and interpretation of median reaction time scores across groups when contrasting control subjects with mTBI subjects. Alternative analyses are presented, including analyses of distribution statistics such as sigma and tau, as well as an analysis which normalizes a subject's reaction time score relative to their population of contributed observations (Z-score transforms within subjects) in order to assess the impact of flankers and cues upon performance while controlling for large differences in base RT and variability across subjects.

These chronometric analyses are then considered within a brain-behavior correlation framework. Diffusion Tensor Imaging analysis of readily identifiable white matter tract structures is used to quantify the presence and severity of damage in mTBI subjects, which are then correlated with reaction time measures associated with mean RT variability, as well as cognitive subtractions isolating conflict, alerting, orienting processes. Results demonstrate significant correlations between degree of white matter tract damage in frontal regions and conflict performance, and further demonstrate separability from both alerting and orienting processes. The relationshipbetween increased RT variability and alerting, orienting, conflict attentional processes will also be discussed.