Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Dissecting activity of individual cell populations with kernel current source density method and independent components analysis

100%

Wojcik D. K., Potworowski J., Glabska H., Leski S.

Acta Neurobiologiae Experimentalis

2012

tom 72

nr 2

Local field potentials (LFP), the low-frequency part of extracellular electric potential, reflect dendritic processing of synaptic inputs to neuronal populations. Today one can easily record simultaneous potentials from multiple contacts. Due to the nature of electric field each electrode may record activity of sources millimeters away which leads to significant correlations between signals and complicates their analysis. Whenever possible it is convenient to estimate the current source density (CSD), the volume density of net transmembrane currents, which generate the LFP. CSD directly reflects the local neural activity and CSD analysis is often used to analyze LFP. We present here a general, nonparametric method for CSD estimation based on kernel techniques, which can take into account known anatomy or physiology of the studied structure. Using data from a simulated large scale model of thalamo-cortical column we also show how CSD analysis combined with independent component analysis (ICA) can reveal information on activity of individual cell populations. Research supported by grants 5428/B/P01/2010/39, POIG.02.03.00- 00-003/09, POIG.02.03.00-00-018/08.

Separating dynamics of different cell populations from multielectrode signals

88%

Wojcik D., Potworowski J., Glabska H., Majka P., Leski S.

Acta Neurobiologiae Experimentalis

2011

tom 71

nr S

Local field potentials (LFP), low-frequency part of extracellular electric potentials, seem to reflect dendritic processing of incoming activity to neural populations. Long-range nature of electric field leads to correlations even between remote recordings showing sources from millimeters away which complicates analysis of LFP. To get more insight it is convenient whenever possible to look for current sources of the potentials or to decompose the signals into meaningful components using statistical techniques. In Łęski and coauthors (2010) we have combined inverse current source density method with independent component analysis (ICA) to decompose 140 recordings in rat forebrain obtaining physiologically meaningful components across a group of seven animals. To find out what can be really observed with such an approach experimentally we simulated local field potentials generated in a single cortical column in a model of 3560 cells with non-trivial morphologies. Having both the current source density (CSD) and LFP generated by twelve cortical populations included we compared it with independent components obtained in the decomposition of data generated by the whole network. We assumed a set of potential measurements on a regular grid, low-pass filtered it temporally under 500 Hz, reconstructed the sources using kernel current source density and performed the ICA. We found that the recorded evoked activity was dominated by two populations of pyramidal neurons, which were well separated by ICA. Other populations could not be clearly distinguished in the simulated potentials nor in the ICA. Supported by grants POIG.02.03.00-00-003/09, POIG.02.03.00-00-018/08.

Ograniczanie wyników

2 Acta Neurobiologiae Experimentalis

2 Glabska H.

2 Leski S.

2 Potworowski J.

1 Majka P.

1 Wojcik D.

1 Wojcik D. K.

1 2012

1 2011

Wyniki wyszukiwania

Dissecting activity of individual cell populations with kernel current source density method and independent components analysis

Separating dynamics of different cell populations from multielectrode signals