One-dimensional CSD analysis requires sampling of LFP profiles using linear array electrodes that fit with some experimental requirements (e.g., the present study), but not with all and several assumptions about the anatomical organization of the brain region to be studied. For these reasons, the first spatial derivative (equivalent to a bipolar recording from closely spaced sites) is a useful alternative (Bollimunta et al., 2008;Ledberg et al., 2007). The first derivative [current flow density (Mitzdorf, 1985)] produces nearly the same attenuation of far-field contamination as the second derivative, but requires only two electrodes. Importantly, the distances and positions of recording electrodes and the choice of differentiation procedure and grid can be determined based on the anticipated generator dimensions (from known anatomy), and can be manipulated experimentally to help define generator properties [see e.g., (Tenke et al., 1993)]. It is noteworthy that use of a bipolar recording is a local solution for the more general “reference electrode problem,” that is of continuing importance in scalp EEG/ERP recordings (Geselowitz, 1998;Nunez et al., 1991;Yuval-Greenberg et al., 2008).
