Abstract

Novelty P3 and frontal response negativity (FRN) are among many EEG/ERP processes (e.g., theta; nogo P3; error-related negativity ERN) that have been attributed to generators within the longitudinal fissure using semiautomatic programs (e.g., BESA) to localize equivalent current dipoles (ECD). These ECDs typically have a radial orientation with a location in the anterior cingulate cortex (ACC) on or near the midline, which is in accordance with results using a less rigid, low resolution inverse method (LORETA). Imaging methods have also been used to determine ECD locations and to validate low resolution findings. Such task- and process-specific findings accord well with the growing literature on the functional significance of midline and basal frontal cortex, and ACC in particular. However, despite what appears to be a parsimonious convergence of findings across methodologies, there is a pervasive problem that is generally unrecognized: inverse models are not intended to account for field potential irregularities caused by local cancellation or field closure. Moreover, the dominant activity expected for a bilateral midline generator should be largely cancelled, and oriented normal to the cortical surface (i.e., tangential to the midline scalp). In contrast to inverse models, the use of more descriptive current source density analysis (CSD; surface Laplacian) methods does not limit field potential irregularities associated with field closure. Using DipoleSimulator, we simulated scalp potentials produced by dipolar generators (unilateral, bilateral, asymmetric) positioned at different depths within the longitudinal fissure (deep: ACC; shallow: supplementary motor area; superficial: bank of fissure), and evaluated LORETA, BESA and CSD solutions. Results for bilateral generators yielded CSD sources that were sharply confined to the frontocentral midline, with the corresponding sinks displaced to the lateral frontal surface, a pattern consistent with previously published CSD maps of novelty P3, and with statistical CSD evidence for FRN. However, sources corresponding to superficial generators yielded sharper topographies, while deep generators were markedly attenuated. Radially-oriented dipoles were also simulated for the medial orbital surface, and for the depths of the cingulate sulcus, resulting in a medial current source that decreased with distance from the midline, but without an abrupt inversion to a current sink. Results indicate that the frontocentral novelty P3 source and FRN sink are generated within the longitudinal fissure, but emphasize the need for additional intracranial CSD evidence.