Talk by Wessel Woldman
- 08 Dec 2011 15:00
- 08 Dec 2011 16:00
A computational model on neuronal stabilization in perceptual choice dynamics
We develop a biophysically realistic network model on the micro-level of perceptual choice dynamics. During intermittently presented visual stimuli, ambiguous stimuli lead to various choice-sequences, such as repetition or alternation. Psychophysical experiments have shown these choice-sequences crucially depend on the duration of blank periods between subsequent stimuli. There exist neurocomputational models at the population level predicting choice sequences adequately, yet the interpretation of certain parameters is unclear. These phenomenological models cannot describe the underlying neuronal mechanisms that take place during these experiments. Neurophysiological experiments by Klink (Klink et al., submitted) provide a framework for understanding the underlying neuronal mechanisms during intermittently presented ambiguous visual stimuli. Klink finds neuronal responses stabilize when off-durations increase and statistical analysis supports the hypothesis of neuronal stabilization. In order to develop a neuronal network that can describe Klink’s data and compare it with psychophysical results, we start from a neuronal network model by Laing and Chow describing binocular rivalry (Laing and Chow 2002). We include on-off-cycles to develop a biophysically realistic neuronal network model that might lead to more insight on the mechanisms of stabilization during repetitive interrupted stimulation with visual stimuli.