Linda Garami, Department of Neuroscience, University of Pennsylvania,
Auditory objects are often detected through regularities in their acoustical structures. Depending on the complexity, these regularities emerge at various timescales. Short and long-term representation based on these regularities contributes to learning and interact with attention that I had been testing in infants, older adults, and older adults with hearing loss using EEG. However, it is hard to tell apart the contribution of different cortical areas and circuits to task performance, which would be crucial to building a valid predictive model that can explain short and long-term learning in the auditory domain. In the study that I am going to present in detail, my goal has been to identify the neuronal circuitry in the auditory cortex (AC) that enables the auditory system to detect temporal patterns in sounds with different complexities.
On a large scale, human data shows that in both the AC and the pre-frontal brain regions differential brain activity correlates with the predictability (random vs. regular) of auditory stimuli. On a smaller scale, responses of individual neurons in AC exhibit sensitivity to temporal regularities in sounds: neurons exhibit stimulus-specific adaptation, a reduction in their response selective to frequently presented inputs. This adaptation may underlie the population sensitivity to more complex spectro-temporal acoustic regularities, and the rapid perception of such regularities.
I adapted a human change-detection paradigm to a rodent model that measures sensitivity to transitions between regular and random acoustic stimuli. Spiking activity of single neurons exhibited stimulus-specific adaptation to the spectral content of the stimuli, that was enhanced in the regular scenes, indicating that the adaptation on the single neuron level in the auditory cortex can evolve at multiple time scales. Increasing the length of the repeating sequence in the predictable environment failed to elicit the same effect, which means that the integration time window might be limited at this level, removing information about previous acoustical contexts relatively rapidly as new stimulus statistics are presented. I will present a simple model of how inhibitory interneurons can play a part in explaining these data. The results underline the importance of the sensory cortex in complex cognitive tasks in the auditory modality.