Source reconstructin methods in analysis of multielectrode recordings
Local field potentials (LFP), the low-frequency part of the extracellular electric potentials, reflect dynamics of the brain at the population level. Because of technological advances it is now feasible to record LFP at tens of locations simultaneously. The interpretation of these signals is complicated by the fact that the electric signals propagate in the tissue, and the signal recorded at each position may have contributions from neurons located more than a millimeter away. Therefore it is useful to estimate and analyze the current source density (CSD), the volume density of transmembrane currents which generate the observed LFP. In the past few years new methods for CSD estimation has been developed, such as the inverse CSD, based on the inversion of the forward-modeling scheme, or the kernel CSD, which employs kernel techniques used in machine learning. I will review these methods and the available software tools.
Rekord w opracowaniu