Network models of frequency modulated sweep detection.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Auditory Processing

Background:

• Frequency modulated (FM) sweeps are crucial in vocalizations, like human speech.
• Auditory neurons exhibit selectivity for FM sweep direction and rate, but underlying synaptic mechanisms are poorly understood.
• Experience-dependent plasticity in FM sweep selectivity remains largely unexplored.

Purpose of the Study:

• To present and analyze three network models of synaptic mechanisms for FM sweep direction and rate selectivity.
• To explain experimental data using these computational models.
• To investigate experience-dependent changes in FM sweep selectivity.

Main Methods:

• Developed three distinct network models: facilitation, duration tuned, and inhibitory sideband.
• Utilized spike-timing-dependent plasticity (STDP) to train the inhibitory sideband model.
• Compared model predictions with experimental data on FM sweep selectivity.

Main Results:

• The facilitation model uses coincidence detection with time-delayed inputs.
• The duration tuned model relies on interactions between delayed excitation and early inhibition.
• The inhibitory sideband model, using frequency-selective inputs, demonstrates direction selectivity trainable via STDP.

Conclusions:

• The presented models offer insights into synaptic and spectrotemporal mechanisms of FM sweep processing.
• These models can explore cellular mechanisms of experience-dependent spectrotemporal processing changes.
• The models have broader applications in studying stimulus movement across sensory epithelia, analogous to visual motion.