Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 2 de 2
Filtrar
Mais filtros










Base de dados
Intervalo de ano de publicação
2.
J Neurosci Methods ; 18(4): 371-83, 1986 Dec.
Artigo em Inglês | MEDLINE | ID: mdl-3796047

RESUMO

Progress in understanding neural information processing in the brain requires the ability to study the correlated activities of neuron groups. However, which neurons belong to a functional group? The answer to this question depends on the availability of multiple electrode arrays and real-time coupling analyzers. This paper describes such an instrument; electrode arrays from our group have been described elsewhere. A real-time covariance computer has been built embodying certain nearly universal neural principles. The spike trains from each recorded unit pass through leaky integrators which mimic the "postsynaptic potentials" (PSPs). The subsequent covariance calculations are performed by multiplications of all possible PSP-channel combinations. This is greatly simplified by the use of analog-to-stochastic converters, which encode the PSPs to stochastic two-level signals. Multiplications, thus, can simply be carried out by AND gates. With a second set of leaky integrators the participations of the single neurons in the joint activity of the group are derived. This operation mimics the spatial and temporal integration at the membranes of the dendritic tree and soma. For each neuron the actual coupling values are displayed in real time via LED bargraph displays. A 4-channel prototype of this instrument was used to analyze covariant activities in the cat's visual cortex.


Assuntos
Simulação por Computador , Modelos Neurológicos , Neurônios/fisiologia , Animais , Gatos , Software
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA
...