Talk abstract:

Specializations in the Auditory Midbrain for Detecting
Temporal Parameters of Biologically Important Sounds

John H. Casseday
Department of Psychology
University of Washington
casseday@u.washington.edu

The auditory midbrain, or inferior colliculus (IC), is an obligatory waystation for >90% of the neurons of the ascending auditory pathways. Thus, nearly all auditory information destined for the thalamus and cortex must pass through synapses in the IC. Cells in the IC receive synaptic input from several different auditory centers in the lower brainstem, each of which produces its own transformation of auditory signals. Therefore, cells in the IC integrate different sorts of auditory information, and from many of them, new signal transformations emerge. Some of the most biologically important of these emergent properties are selective responses to time varying stimuli. These properties emerge from the interaction of excitatory and inhibitory inputs, temporally offset from one another. We have studied two kinds of selectivity in the IC of the big brown bat-neural tuning for sound duration and tuning for the rate of repetitive frequency modulations. Duration tuning refers to the ability of some neurons to respond to a narrow range of sound duration but not to shorter or longer sounds. For example, a duration tuned neuron may give a vigorous response to a sound 5 ms long but completely fail to respond to sound <3 ms or >7 ms. This neuron has a best duration of 5 ms; different duration tuned neurons have different best durations. A fairly simple model, comprised of three inputs, can account for duration tuning: 1) Leading inhibition--inhibitory input has the shortest latency and arrives prior to the excitatory input; it is a sustained response that lasts at least as long as the duration of the sound. 2) Offset excitation--an excitatory event is correlated with the offset of sound; its magnitude is below the threshold for spike initiation. 3) Delayed excitation--excitatory and transient input has a long latency relative to the onset of sound; it is temporally locked to the onset of sound; it arrives after the initial inhibitory input; its magnitude is just at or below spike threshold. The neuron fires when the two excitatory inputs coincide in time with one another. The best duration of a neuron is determined by the time delay between the onset of inhibitory input and the delayed excitation. Other specialized neurons respond to repetitive frequency modulations but do not respond to simple sounds such as pure tones or noises. Moreover, these neurons are tuned for the rate of frequency modulations. This sort of specialization can be accounted for by a modification of the duration tuning model. The most important modification is that the excitatory and inhibitory components are linked to different frequency components of the sound. These two types of neurons appear to be filters for the same range of temporal parameters of sounds as encountered by the bat during detection of prey or avoidance of predators.

Research supported by NIH grants DC-00287 and DC-00607.

Back to Workshop Schedule

1998-1999 Mathematics in Biology