Direct and indirect integration of event-related potentials, functional magnetic resonance images, and single-unit recordings.

Cognitive neuroimaging techniques vary along three primary dimensions: invasiveness, temporal resolution, and spatial resolution. Several of the major techniques excel on two of these three dimensions, but none of them excels on all three. In principle, multiple techniques with different strengths and weaknesses could be combined to obtain high temporal and spatial resolution data about human neural activity, and this article compares two approaches to combining microelectrode, hemodynamic, and electromagnetic measures of neural activity. The first approach involves using structural magnetic resonance images to provide a common reference frame for the mathematical estimation of neural activity, and the second approach involves parallel experimental manipulations and converging evidence. At present, neither approach is entirely satisfactory, and the integration of different measures of neural activity, therefore, requires a combination of direct and indirect approaches.