Behavioural and neural correlates of vibrotactile discrimination and uncertainty

Abstract

Decisions pervade everyday life, from the mundane to those that induce anxiety. The act of making simple decisions may be examined experimentally with the vibrotactile discrimination task, where the manipulation of task factors challenge the way people make decisions. The objective of this thesis is to examine how the human brain maintains above-chance levels of performance despite challenging experimental conditions that alter the way people perceive presented sensory stimuli. Two behavioural investigations of perceptual decision making and two separate analyses of a functional neuroimaging experiment were conducted. In the first experiment, we examine the influence of the time-order effect whereby prior information from task stimuli biases decision making on a current trial. In the second study, different delay periods between vibration pairs were used to examine how the working memory representation of a vibrotactile stimulus drifts over time by observing changes in accuracy and sensitivity. The neural correlates of explicit factors used in the task, including stimulus noise and context judgements, were then studied through a functional neuroimaging experiment. A number of distinct prefrontal neural regions were identified, a selection of which were then used in the model-driven network based technique of dynamic causal modelling. This thesis makes the following conclusions: Even when not explicitly incorporated into experimental design, the history of previously presented stimuli can quickly establish an internal standard and exert a powerful influence on decision making. The time-order effect exhibits its influence on decision making in a nonlinear fashion across short interval delay periods between paired stimuli, in a way that depends upon prior experience with time-dependent tasks. Distinct prefrontal cortex regions including the inferior frontal gyrus pars triangularis and the superior frontal gyrus, are engaged when precision estimates of stimulus representations are required for decision making. These prefrontal regions exert their influence through nonlinear, hierarchical network connections. The findings of this thesis could be extended to elucidate cognitive disturbances in depression where deficits in decision making are a debilitating daily experience.