In cognitive neuroscience, neuroimaging methods are used in order to visualize brain activity associated with cognitive processes in time and space. Modern techniques like PET (Positron Emission Tomography) and fMRI (functional Magnetic Resonance Imaging) offer the possibility to accurately localize changes in regional cerebral blood flow. However, the temporal resolution of these techniques is rather poor, since changes in blood flow are much slower (in the order of several seconds) than the neural activity eliciting these changes. Another neuroimaging method is source analysis on the basis of event-related potentials (ERPs). ERPs are obtained by averaging epochs of the electroencephalogram (EEG) time-locked to the occurrence of concrete sensory, motor, or cognitive events. ERPs reflect that part of the brain activity that is specifically related to the processing of these events. With ERPs it is possible to study information processing in the brain with a high temporal resolution (ms or less). Because of this it is possible to use sophisticated experimental designs in which stimulus and/or task parameters are varied on a trial-by-trial basis and to search for subtle relations between task performance and ERP components. For these reasons, ERPs are a powerful tool for studying the organization and time course of elementary processes in selective attention, memory, language, motor control, etc.

However, the relationship between the scalp-recorded brain activity and the underlying neural activity is complex and not completely understood. A common approach is to attempt to infer the activated sources in the brain on the basis of multi-channel ERP recordings and inverse dipole modeling. However, in this approach several basic assumptions have to be made regarding the nature of the volume conductor (i.e. the model of the head) and the sources of electrical activity in the brain. In the spatio-temporal dipole model [2], for instance, it is assumed that a certain latency range of ERP activity is explained by a small number of equivalent dipoles, the locations and orientations of which remain fixed over time. In complex cognitive tasks, however, ERPs may reflect many overlapping source activities, complicating dipole modeling. I will discuss an approach in which topographical analysis and dipole modeling is restricted to the lateralized part of the brain activity associated with covert orienting of attention and preparation of eye movements. The approach rests on a double subtraction procedure in which ipsilateral ERPs are subtracted from contralateral ERPs and averaged over left and right motor responses (LRPs [1]; in this case: attention shifts and eye movements). This approach was originally developed for ERP derivations from a single electrode pair (C3/C4), and now applied to multi-channel topographical analyses. Using this approach, all non-lateralized brain activity is removed from the analysis, and thus it is more likely that relatively simple dipole models will suffice. Indeed it was found that this procedure resulted in relatively simple topographies in which an occipital asymmetry (probably related to attentional orienting) preceded a frontal asymmetry (probably related to oculomotor preparation). Furthermore, these results were highly consistent over different individual subjects.

References

1. Jong, R. de; Wierda, M.; Mulder, G.; Mulder, L.J.M. (1988). The use of partial information in response preparation. Journal of Experimental Psychology: Human Perception and Performance, 14, 682-692.
2. Scherg, M. (1990). Fundamentals of dipole source potential analysis. In: F. Grandori, M. Hoke & G.L. Romani (Eds.), Auditory Evoked Magnetic Fields and Potentials. Advances in Audiology, 6, 40-69. Basel: Karger.