Functional neuroimaging of dual task interference and divided attention

Abstract: It is known that human subjects cannot perform two reaction time (RT) tasks as efficiently as they would perform the constituent tasks individually. Increments of reaction times or error rates that occur when two RT tasks are performed nearly simultaneously is known as interference. The neurophysiological mechanism or mechanisms behind this phenomenon has remained largely unexplored. This thesis describes a series of experiments designed to investigate psychophysical and neurophysiological aspects of interference. We psychophysically studied mechanisms behind how simple reaction times are prolonged when they are performed for stimuli presented with short inter-stimulus intervals. These psychophysical experiments were later examined with functional neuroimaging for putative neurophysiological mechanisms of interference, namely, occupancy of motor structures by a task, division of attention between sensory modalities, and effector assignment strategies. Thus, our findings are presented in a neurobiological framework to explain interference, in contrast to the previously existed psychological models. We found that simple reaction time tasks, when performed in close succession, produce interference of several different patterns. Firstly, the reaction time to the second of a pair of stimuli increases linearly as a function of the inter-stimulus interval when it is below 400 ms. This RT increase is psychophysically separable from the increment of RT to the first of the stimuli pair, which occurs only when subjects divide attention between two sensory modalities. The increment in RT to the second stimulus could be due to strategies that the nervous system uses to specify effectors, in that when the effector can only be specified a posteriori to stimuli, there is a longer increment of RT than when the effector is specified a priori. Corresponding to these psychophysical findings, we found that the motor tasks activate a nearly identical spatial extent in motor structures that include the primary motor cortex, supplementary and cingulate motor areas, basal ganglia and the ventral lateral and ventral anterior thalamic nuclei. When two reaction time tasks requiring motor control of two effectors must be near simultaneously controlled by a large proportion of neurons in these structures, this cannot be done, because the control of the effector to the first stimulus appears to occupy these neuronal populations for a finite length of time. When this happens, the response to the second stimulus is delayed, which is the behavioral outcome that indicates interference. We have demonstrated that the occupancy of motor structures lead to the activation of an additional cortical area located in the right inferior frontal gyrus, and that, this activity is strongly correlated to the delay in the RT to the second stimulus. In addition, we found that division of attention prolongs the RT to the first stimulus and that this activates a set of brain regions that are located in the caudal superior frontal areas and intra-parietal areas bilaterally. These areas are anatomically dissociable from the activity in the motor structures. The findings of these experiments have raised the need to reconcile the notion of computational parallel processing in the brain with the interference phenomena that occur under concurrent multitasking situations. Investigating these phenomena with specific neurophysiological hypotheses would, on the one hand, allow us to elucidate why the human brain is limited in certain aspects of its information processing capabilities, and on the other, provide organizing principles to understand how the brain works.