Berlin 2012 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

BP: Fachverband Biologische Physik

BP 18: Posters: Statistical Physics in Biological Systems

BP 18.34: Poster

Mittwoch, 28. März 2012, 17:30–19:30, Poster A

Describing non-linear attentional modulation patterns through ring-architecture neural circuits — •Markus Helmer^1,4, Vladislav Kozyrev^2,3, Stefan Treue^2,4, Theo Geisel^1,4, and Demian Battaglia^1,4 — ¹Max Planck Institute for Dynamics and Self-Organization, Göttingen — ²Deutsches Primatenzentrum, Göttingen — ³Institute of Neuroinformatics, Ruhr-University Bochum — ⁴BCCN, Göttingen

We analyzed recordings in area MT from macaque monkeys performing a transparent motion task. They were presented random-dot-patterns (RDP) at two distinct locations within the receptive field of the recorded cell. Attention was directed to a fixation spot or to only one of the two RDPs. The angle between the two RDPs was kept fixed at 120 degrees so that covarying the motion directions provided tuning curves with two peaks. We found that the positions of the response peaks were different from what the single-motion orientation-preferences of the cell predicted. Furthermore they depended on the attentional condition, showing strong signatures of non-linear interactions during the integration of the two stimuli by MT neurons, including peak repulsion or attraction. By using a mean-field ring model, we could reproduce all the observed response non-linearities independently in broad regions of the parameter space. However, to find the specific combination of non-linearities observed across attentional conditions a fine tuning of the parameters was required. We explored therefore multi-areal network models with multiple coupled rings to achieve a more robust description of non-linear attentional modulation patterns.