Biophysical Model of Coincidence Detection in Single Nucleus Laminaris Neurons
Abstract
A new computational model of a Nucleus Laminaris neuron investigates the cell's ability to detect coincides between ipsilateral and contralateral inputs (arriving from Nucleus Magnocellularis). The model, constructed in the NEURON programming environment, is biophysically detailed. The role of dendritic processing is explored and emphasized, including the effects of active potassium channels in the dendrites, dendrite size on post-synaptic potentials (PSPs) synaptic reversal potentials on PSPs, synaptic distribution along the dendrites, parallel (multiple) dendritic processing, and the benefits of having dendrites at all. The model makes several predictions, some which confirm those of a previous model (Agmon-Snir et al., 1998). Active potassium channels greatly increase the ability to detect coincidences at high frequencies. Rate-coded output is much more robust than vector-strength-coded output from partial coincidences. The phase locking of the output spikes is sharper than the phase locking of the synaptic inputs. Densities of active potassium channels are significant, but substantially smaller than those found experimentally in vestibular neurons. There is an optimal number of dendrites for a given stimulus frequency, dendritic length, and number of synaptic inputs. Ceiling effects from the synaptic reversal potential decreases the rate of false detections. The most efficient dendritic length decreases with best frequency. [Simon, J. et al. 1999, Assoc. Res. Otolaryngol. Abs: 572.]
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 16, 1999
- Accession Number
- ADA484247
Entities
People
- Catherine E. Carr
- Jonathan Z. Simon
- Shihab A. Shamma
Organizations
- University of Maryland