Dynamical modeling and cortical oscillations in Serinus canaria song

S. BOARI, G.B. MINDLIN , A. AMADOR

Conferencia; SAN 2020 XXXV Annual (Virtual) Meeting; 2020

Sociedad Argentina de Neurociencias

Sensorimotor integration is a crucial aspect for encoding auditory representations of learned vocal behavior. Songbirds provide an excellent animal model to address this, since they are vocal learners and the neural circuits for song learning, production and perception are well understood. A dynamical model for the population activity in brain regions of this ?song system? can reproduce characteristic air sac pressure patterns of canary song. For a cortical sensorimotor nucleus that presents similar activity patterns while singing and while perceiving the bird?s own song (BOS), the model predicts synchronized population activity at distinct temporal instances of song syllables. In this talk, I will discuss the neural model and its predictions. Also, I will report evidence of synchronous activity in response to auditory playbacks of the BOS. Extracellular recordings using 32-channel silicon probes allowed to study the local field potentials (LFP), single unit activity, and spatial synchronization. Rhythmic features of the song of canaries (Serinus canaria) allowed us to uncover neural oscillations locked to the behavior, which shows that predictions are compatible with experiments. Grouped data from different subjects share these features, suggesting a general strategy for neural coding. This program illustrates an example of the power that dynamical models provide to neuroscience research, in the form of testable hypotheses that drive -and feedback from- experiments.