Integrate-and-fire-type models of the lateral superior olive.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Auditory System Modeling

Background:

• Neurons in the lateral superior olive (LSO) are crucial for binaural sound localization.
• Existing LSO neuronal models are often specific and their interrelationships and applicability are unclear.

Purpose of the Study:

• To compare six single-compartment integrate-and-fire (IF) LSO models.
• To determine model suitability for different computational neuroscience applications.
• To evaluate the impact of subthreshold properties and spike generation mechanisms on model performance.

Main Methods:

• Simulated six IF LSO models with identical synaptic inputs.
• Classified models into passive (leak conductance) and active (potassium conductance) groups.
• Subdivided models based on spike generation mechanisms (simple, spike shape mimicry, exponential current).
• Calibrated models using common binaural tuning criteria.

Main Results:

• Active models showed higher firing rates for intense inputs and lower rates for temporal inputs compared to passive models.
• Responses within passive or active groups were similar across different spike generation types.
• Active IF models demonstrated superior accuracy in reproducing temporal coding of LSO neurons.

Conclusions:

• Active models incorporating potassium conductance are more suitable for accurate temporal coding in LSO models.
• The choice of spike generation mechanism has less impact on overall response patterns than subthreshold properties.
• Extended spiking mechanisms for realistic spike shapes incur minimal computational cost.