An Interneuron Circuit Reproducing Essential Spectral Features of Field Potentials

作者：Maex Reinoud^*

来源：Neural Computation, 2018, 30(5): 1296-1322.

DOI：10.1162/neco_a_01068

Recent advances in engineering and signal processing have renewed the interest in invasive and surface brain recordings, yet many features of cortical field potentials remain incompletely understood. In the computational study that follows, we show that a model circuit of interneurons, coupled via both GABA(A) receptor synapses and electrical synapses, reproduces many essential features of the power spectrum of local field potential (LFP) recordings, such as 1/f power scaling at low frequency (below 10 Hz), power accumulation in the -frequency band (30-100 Hz), and a robust rhythm in the absence of stimulation. The low-frequency 1/f power scaling depends on strong reciprocal inhibition, whereas the rhythm is generated by electrical coupling of intrinsically active neurons. As in previous studies, the power arises through the amplification of single-neuron spectral properties, owing to the refractory period, by parameters that favor neuronal synchrony, such as delayed inhibition. This study also confirms that both synaptic and voltage-gated membrane currents contribute substantially to the LFP and that high-frequency signals such as action potentials quickly taper off with distance. Given the ubiquity of electrically coupled interneuron circuits in the mammalian brain, they may be major determinants of the recorded potentials.

• 出版日期2018-5