Abstract
Attention to visual features increases the firing rate for neurons tuned to those features (e.g. Saenz, Buracas & Boynton, 2002). In this study, we measure the effects of attention to point-light biological motion on neural activity in human STS, a brain area believed to have neural populations tuned to distinct actions (e.g. Grossman, Jardine & Pyles, 2010). Methods. Observers viewed a 1 sec animation of a centrally positioned point-light walker. A 50ms square appeared 3deg in the periphery at 300ms or 600 ms after the onset of the walker. In different blocks subjects were instructed to report 1) the facing direction of the walker (attend to walker, ignore flanker), 2) the position of the flanker (ignore walker, attend to flanker), or 3) the facing direction of the walker, but hold the response until the flanker appeared (attend to both). Results. Principal component analysis on the STS BOLD response revealed three spatially distinct subdivisions within the independently localized STS. Neural activity in the fundus of the STS was strongest when subjects attended to the biological motion, while the most dorsal aspects of the STS were more activated when monitoring for the peripheral flanker. The most anterior regions of the STS deactivated during all three attention tasks. Functional connectivity analyses revealed these regions to be correlated with homologous regions in the opposite hemispheres, but not with each other (even within the same hemisphere). Conclusions. Our findings reveal a coarse cortical organization in the STS during action recognition. The increased neural activity in the STS fundus likely reflects a gain in action selective neurons as a result of attending to the biological motion. The other subdivisions of the STS likely reflect the domain-general attentive demands of perceptually organizing biological motion.
Meeting abstract presented at VSS 2013