Kernel Current Source Density method. CSD estimation for arbitrary distribution of contacts in one, two, and three dimensions.
-
1
Nencki Institute of Experimental Biology, Department of Neurophysiology, Poland
-
2
WLOG Solutions, Poland
Extracellular recordings of electric potential are an important technique in the studies of neural activity in vivo and in vitro. In the last few years we observe rapid development of technology for large scale electrical recordings. Various types of multi-electrodes were devised to simultaneously record extracellular potentials from multiple spatial locations. The low-frequency part of these recordings, the local field potentials (LFP), is considered to be a signature of the dendritic processing of synaptic inputs. Since LFP is a non local measure of neural activity, with contributions from neurons located more than a millimeter away from the electrode, its direct interpretation is difficult. Thus if only possible it is useful to estimate the current source density (CSD), the volume density of net transmembrane currents, which is the source of the LFP. CSD is directly related to the local neural activity and current source density analysis is a popular tool in the analysis of multivariate LFP.
In homogeneous and isotropic tissue CSD is given by the laplacian of the potentials, so discrete differentiation is the simplest estimate if we have a set of potentials measured on a regular grid. However, if we interpolate so obtained values and ask if the sources which generated the potentials we measured can be given by the interpolated CSD, the answer is usually negative. To find a better answer to the question of possible generating source recently a new method for CSD estimation has been developed, called the inverse CSD (iCSD) method. The main idea behind iCSD is to assume a specific parametric form of CSD generating potentials, calculate the LFP in a forward-modeling scheme to obtain the values of CSD parameters. The iCSD framework developed so far requires an assumption of a specific geometry of contacts and new calculations are needed for every new electrode distribution. All the variants up to now assumed recordings on regular, rectangular grids. Moreover, the complexity of the reconstructed CSD distribution was limited by the number of observations.
Here we present a new, nonparametric method for CSD estimation. The kernel CSD method (kCSD) is based on kernel techniques, widely used in machine learning. kCSD lets the user specify the family of allowed CSD distributions in more intuitive way through over-complete bases. It also makes it possible to employ knowledge of anatomy and physiology of the probed structure, such as laminar structure. The assumption of regular electrode arrangement is not necessary, we show how kCSD can be applied to recordings from electrodes distributed at any positions on one-, two-, and three-dimensional sets with equal ease. Moreover, it turns out that kCSD is a general non-parametric framework for CSD estimation including all the previous variants of iCSD methods as special cases.
Partly supported from grants POIG.02.03.00-00-003/09 and PBZ/MNiSW/07/2006/11.
References
1. Pettersen, KH, et al. (2006) Current-source density estimation based on inversion of electrostatic forward solution: effects of finite extent of neuronal activity and conductivity discontinuities. J Neurosci Meth 154: 116-33.
2. Wirth, C, Luscher, HR (2004) Spatiotemporal evolution of excitation and inhibition in the rat barrel cortex investigated with multielectrode arrays. J Neurophysiol 91: 1635.
3. Łęski, Szymon, et al. (2007) Inverse Current-Source Density Method in 3D: Reconstruction Fidelity, Boundary Effects, and Influence of Distant Sources. Neuroinformatics 5: 207-222.
Conference:
Neuroinformatics 2010 , Kobe, Japan, 30 Aug - 1 Sep, 2010.
Presentation Type:
Poster Presentation
Topic:
Electrophysiology
Citation:
Potworowski
J,
Jakuczun
W,
Łęski
S and
Wòjcik
D
(2010). Kernel Current Source Density method. CSD estimation for arbitrary distribution of contacts in one, two, and three dimensions..
Front. Neurosci.
Conference Abstract:
Neuroinformatics 2010 .
doi: 10.3389/conf.fnins.2010.13.00114
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
15 Jun 2010;
Published Online:
15 Jun 2010.
*
Correspondence:
Daniel Wòjcik, Nencki Institute of Experimental Biology, Department of Neurophysiology, Warszawa, Poland, d.wojcik@nencki.edu.pl