Detecting Coherence in Neuronal Data

摘要: Spatially and temporally coherent activities emerge in many neuronal systems. The analysis of such responses most cases is based on simple correlation techniques which cannot detect nonlinear relationships. In this contribution I review new approaches for the signals. On one hand present model free are general notion statistical dependency apply to neurophysiological observables spike local field potentials, respectively. These quantify coherence information theoretical terms can help characterize underlying dynamics. show that contributions be analyzed a time resolved way method allows identification episodes background stochastic activity. other model-dependent approach particularly well suited assemblies exhibiting emergent burst activities. It assumes form network dynamics parameters directly determined from experimental trains. This Ansatz deals with fact observable trains only stochastically reflect an Despite mathematical simplicity it determines important characteristics like memory switching behavior synchronization methods illustrated by multiunit activity potential data visual cortex cat. Both independently reveal these switches between essentially two states, oscillatory one. within oscillations synchronizations occur identified high resolution either potentials as turns out across cortical distance quite rare event occurs only.