RT Journal Article
SR Electronic
T1 Brain Bases for Auditory Stimulus-Driven Figure–Ground Segregation
JF The Journal of Neuroscience
JO J. Neurosci.
FD Society for Neuroscience
SP 164
OP 171
DO 10.1523/JNEUROSCI.3788-10.2011
VO 31
IS 1
A1 Sundeep Teki
A1 Maria Chait
A1 Sukhbinder Kumar
A1 Katharina von Kriegstein
A1 Timothy D. Griffiths
YR 2011
UL http://www.jneurosci.org/content/31/1/164.abstract
AB Auditory figure–ground segregation, listeners' ability to selectively hear out a sound of interest from a background of competing sounds, is a fundamental aspect of scene analysis. In contrast to the disordered acoustic environment we experience during everyday listening, most studies of auditory segregation have used relatively simple, temporally regular signals. We developed a new figure–ground stimulus that incorporates stochastic variation of the figure and background that captures the rich spectrotemporal complexity of natural acoustic scenes. Figure and background signals overlap in spectrotemporal space, but vary in the statistics of fluctuation, such that the only way to extract the figure is by integrating the patterns over time and frequency. Our behavioral results demonstrate that human listeners are remarkably sensitive to the appearance of such figures. In a functional magnetic resonance imaging experiment, aimed at investigating preattentive, stimulus-driven, auditory segregation mechanisms, naive subjects listened to these stimuli while performing an irrelevant task. Results demonstrate significant activations in the intraparietal sulcus (IPS) and the superior temporal sulcus related to bottom-up, stimulus-driven figure–ground decomposition. We did not observe any significant activation in the primary auditory cortex. Our results support a role for automatic, bottom-up mechanisms in the IPS in mediating stimulus-driven, auditory figure–ground segregation, which is consistent with accumulating evidence implicating the IPS in structuring sensory input and perceptual organization.