Greg Conradi Smith

Greg Conradi Smith

biology as a mathematical science, science among the liberal arts

Visual Thalamus

Joint work with John Rinzel, Lee Cox, and S. Murray Sherman.

This past computational neuroscience project developed conductance-based network models and population-density models of the lateral geniculate nucleus, a visually responsive region of the thalamus. The central biological question was how inhibitory circuitry associated with the dorsal lateral geniculate nucleus (dLGN) and perigeniculate nucleus (PGN) shapes visual relay.

Thalamocortical Relay

The project began with a minimal integrate-and-fire-or-burst model of thalamocortical relay neurons. The model reproduced burst and tonic response properties observed during sinusoidal current injection in cat thalamic slice preparations containing the dLGN and PGN.

Because the model was compact, it could be constrained using Fourier analysis of current-clamp recordings. Subsequent work added spike-frequency adaptation and used electrophysiological protocols to quantify adaptation time constants and firing-rate reductions in thalamocortical neurons.

Network Dynamics

The modeling framework was used to study mode-locking, detectability of excitatory versus inhibitory input, feedback inhibition, and throughput in retinogeniculate transmission. These studies connected single-cell response properties to network-level questions about how retinal input is relayed through the thalamus.

Population Density Models

A later phase of the project developed two-dimensional population-density models of the dLGN/PGN relay. These models represented the distribution of membrane potential and low-threshold Ca2+ current de-inactivation across large neuronal populations, making it possible to compare population-density calculations with Monte Carlo simulations of corresponding integrate-and-fire-or-burst networks.

The population-density approach reproduced rhythmic bursting in the absence of retinal input and a range of aroused responses under simulated neuromodulation and retinal drive. It provided a substantially more efficient alternative to large Monte Carlo network simulations while preserving the key stochastic and cellular-state variables.

Selected Publications

  • Smith GD, Cox CL, Sherman SM, and Rinzel J. Fourier analysis of sinusoidally driven thalamocortical relay neurons and a minimal integrate-and-fire-or-burst model. J. Neurophys. 83(1):588-610, 2000. [journal] [PMID:10634897]
  • Smith GD, Cox CL, Sherman SM, and Rinzel J. A firing-rate model of spike-frequency adaptation in sinusoidally-driven thalamocortical relay neurons. Thalamus and Related Systems 1(2):135-156, 2001. [doi:10.1016/S1472-9288(01)00012-7]
  • Coombes S, Owen MR, and Smith GD. Mode-locking in a periodically forced integrate-and-fire-or-burst neuron model. Phys. Rev. E 64(041914):1-12, 2001. [doi:10.1103/PhysRevE.64.041914] [PMID:11690059]
  • Smith GD and Sherman SM. Detectability of excitatory vs. inhibitory drive in a thalamocortical relay neuron model. J. Neurosci. 22(23):10242-10250, 2002. [PMID:12451125]
  • Huertas MA, Groff JR, and Smith GD. Feedback inhibition and throughput properties of an integrate-and-fire-or-burst network model of retinogeniculate transmission. J. Comput. Neurosci. 19(2):147-180, 2005. [doi:10.1007/s10827-005-1084-6] [PMID:16133817]
  • Huertas MA and Smith GD. A multivariate population density model of the dLGN/PGN relay. J. Comput. Neurosci. 21(2):171-89, 2006. [doi:10.1529/biophysj.105.075036] [PMID:16788765]
  • Huertas MA and Smith GD. Population density model of the driven LGN/PGN. In: Stochastic Methods in Neuroscience. Laing C and Gabriel L, eds. Pages 217-241. Oxford University Press. 2009. [doi:10.1093/acprof:oso/9780199235070.003.0008]