EducationPhD, Johns Hopkins University School of Medicine (2008)
We hear sounds such as speech in a wide range of listening conditions, of which few parameters are under our control. For example, sounds might emanate from different locations, at different intensities, in the presence of other noise or distracting sounds, and in echoing settings. For a given sound, each of these environmental variables alters the physical pressure waveform that impinges on our eardrums; yet, we are able to interpret these varied physical waveforms as arising from the same underlying sound. In other words, our perception of sounds is invariant to a large number of nuisance parameters, and one of the primary functions of the ascending auditory system is to develop this perceptual invariance. The research thrust of our laboratory is to study the mechanisms by which such invariance properties are generated in the neural responses of primary and higher auditory cortex. In particular, we focus on one behaviorally important set of sounds - vocal communication sounds.
Understanding how the brain processes sounds in realistic conditions is a central problem in auditory neuroscience. The best example of the impressive human ability to 'tune in' on particular sounds in noisy environments is the “cocktail party effect” – where a listener in a crowded, loquacious room can attend to one particular voice of interest. This ability is unmatched by modern speech recognition algorithms, which have accurate performance in silence, but greatly degraded performance in such real-world situations. Yet, the mechanisms and computations by which the brain accomplishes this feat is largely unknown. We hope to answer the fundamental questions of what computations the brain might be using to solve this problem, and what underlying circuitry supports these computations.
We use a range of techniques to answer these questions – in-vivo array and multi-electrode extracellular recordings from awake (and eventually behaving) animals, in-vivo whole-cell intracellular recordings, pharmacological manipulations of neural activity and computational modeling. We expect to add in-vivo two-photon imaging, inducible genetics and viral tract-tracing to this suite of techniques in the near future.
Realistic listening conditions pose a significant challenge to patients with communication disorders such as dyslexia, some sensory aphasias, and to the hearing impaired. We hope to provide fundamental insights into these disorders by understanding the circuit mechanisms by which the brain extracts meaningful signals from noise.
Sadagopan, S., Temiz-Karayol, N.Z. and Voss, H.U. High-field functional magnetic resonance imaging of vocalization processing in marmosets. BioRxiv doi: http://dx.doi.org/10.1101/010561, 2014.
Sadagopan, S. and Ferster, D. Feedforward origins of response variability underlying contrast invariant orientation tuning in cat visual cortex. Neuron 74: 911 – 23, 2012.
Sadagopan, S. and Wang, X. Contribution of inhibition to stimulus selectivity in the primary auditory cortex of awake primates. The Journal of Neuroscience 30: 7314 – 25, 2010.
Sadagopan, S. and Wang, X. Nonlinear receptive fields underlie feature selectivity in primary auditory cortex. The Journal of Neuroscience 29: 11192 – 202, 2009.
Sadagopan, S. and Wang, X. Level invariant representation of sounds by populations of neurons in primary auditory cortex. The Journal of Neuroscience 28: 3415 – 26, 2008.