Physics Colloquium - Monday, Feb. 2nd, 2009, 4:00 P.M.

Woodrow Shew
National Institutes of Health

Recent experiments demonstrate that the activity of neurons in the brain has statistical properties expected near the critical point of a phase transition. These observations raise the question; why might the brain operate at such a critical point? Theories and models predict optimized information processing, but experimental support for these ideas has been lacking.

Here, we experimentally test the prediction that the dynamic range, i.e. the range of stimulus intensities the neural network can process, is greatest at a critical point. We study brain tissue cultures and compare our results to a simple numerical model. We tuned the system through a phase transition using drugs that alter the relative influence of excitatory and inhibitory interactions between neurons. In agreement with our model, the critical point of the phase transition corresponds to the situation where excitatory and inhibitory interactions balance each other. We find that dynamic range is indeed maximized when the neural tissue is closest to the phase transition.