Searching for Temporal Phase Coherence in the Cortical Network with a Time-Frequency Analysis of the Local Field Potential

Gamma-band (25-90Hz) oscillations occur in many parts of the brain. We are seeking to understand the underlying neural mechanisms that generate gamma oscillations in the cerebral cortex by studying gamma activity in the Local Field Potential (LFP) in V1 cortex. The LFP is a measure of the average membrane potential of neurons in the neighborhood of the recording electrode and is commonly interpreted as a measure of local network activity (Kruse and Eckhorn 1996; Logothetis et al. 2001; Buzsaki 2006). Previous experimental studies of V1 have reported a peak in the gamma-band of he LFP power spectrum when the visual cortex was visually driven. (Gray and Singer 1989; Frien et al. 2000; Logothetis et al 2001; Siegel and Konig 2003; Henrie and Shapley 2005). In modeling studies of neuronal networks of inhibitory neurons, and with both inhibitory and excitatory neurons, theorists have found that neurons in these networks may exhibit synchronized spiking that gives rise to deterministic, harmonic network oscillations (Abbott and van Vreeswijk 1993; Somers and Kopell 1993; Wang and Rinzel 1993; Somers and Kopell 1995; Wang and Buzsaki 1996; Traub et al. 1996a; Traub et al. 1996b; Ermentrout and Kopell 1998; White el al. 1998; Kopell et al. 2000; Maex and De Schutter 2003; Borgers and Kopell 2005; Borgers et al. 2005; Tuckwell 2006, Kumar 2008). Several other modeling studies (Brunel and Hakim 1999; Brunel 2000; Tiesinga and Jose 2000; Brunel and Wang 2003; Geisler et al. 2005; Brunel and Hansel 2006) have found that there are parameter regimes in which individual cells do not exhibit periodic firing but the network-averaged firing rate does, and moreover exhibits a temporally coherent, periodic oscillation in the gamma-band. These studies find that the networks in these studies have a harmonic or relaxation oscillator response in their subthreshold membrane potential, and noise in the network causes individual neurons to fire preferentially at the peaks of the membrane oscillation. A common feature of these models is that the coherent network oscillation is an emergent behavior of the system and is an equilibrium or steady dynamical state of the network as long as the statistics of the network inputs remain constant. That such temporally coherent oscillations underlie the experimentally observed gamma-band peak we refer to as the coherent oscillator hypothesis. The coherent oscillator hypothesis, in the form of either spike or membrane potential synchrony, has been widely cited in the experimental literature on gamma-band oscillations but the temporal coherence of these oscillations has not been examined statistically. The concept of emergent coherent oscillations appears in studies of the hippocampus in vivo and in vitro (Penttonen et al. 1998; Fisahn et al. 1998; Traub et al. 2000; LeBeau et al. 2002; Mikkonen et al. 2002; Csicvari et al. 2003; Mann et al. 2005a; Mann et al. 2005b; Vida et al. 2006; Mann and Paulsen 2007; Montgomery and Buzsaki 2007; Senior et al. 2008), visual cortex (Henrie and Shapley 2005, Zaksas and Pasternak 2006; Disney and Aoki 2008), neocortex (Giugliano et al. 2004; Morita et al. 2008), prefrontal cortex (Compte et al. 2003; Durstewtiz and Gabriel 2006), olfactory systems of insects and mammals (Macleod and Laurent 1996; Macleod et al. 1998; Lagier et al. 2004; Bathellier et al. 2006; Schoppa 2006) and other regions of the brain (Buhl et al. 1998; Edwards et al. 2005; Sohal and Huguenard 2005; Joelving et al. 2007). In the EEG literature there have been studies of the phase-locking between visual stimulus and the gamma-band response in EEG signals recorded in humans (Tallon-Baudry et al. 1996; Frund et al. 2007; Senkowski et al. 2007, Schadow et al. 2007, Ohla et al. 2007) Here we analyze LFP measurements recorded from macaque V1. Visual stimulation evokes a noisy re-