Temporal binding of neural responses for focused attention in biosonar

Abstract

Big brown bats emit biosonar sounds and perceive their surroundings from the delays of echoes received by the ears. Broadcasts are frequency modulated (FM) and contain two prominent harmonics sweeping from 50 to 25 kHz (FM1) and from 100 to 50 kHz (FM2). Individual frequencies in each broadcast and each echo evoke single-spike auditory responses. Echo delay is encoded by the time elapsed between volleys of responses to broadcasts and volleys of responses to echoes. If echoes have the same spectrum as broadcasts, the volley of neural responses to FM1 and FM2 is internally synchronized for each sound, which leads to sharply focused delay images. Because of amplitude–latency trading, disruption of response synchrony within the volleys occurs if the echoes are lowpass filtered, leading to blurred, defocused delay images. This effect is consistent with the temporal binding hypothesis for perceptual image formation. Bats perform inexplicably well in cluttered surroundings where echoes from off-side objects ought to cause masking. Off-side echoes are lowpass filtered because of the shape of the broadcast beam, and they evoke desynchronized auditory responses. The resulting defocused images of clutter do not mask perception of focused images for targets. Neural response synchronization may select a target to be the focus of attention, while desynchronization may impose inattention on the surroundings by defocusing perception of clutter. The formation of focused biosonar images from synchronized neural responses, and the defocusing that occurs with disruption of synchrony, quantitatively demonstrates how temporal binding may control attention and bring a perceptual object into existence.