|
Stochastic dynamics, chaos or self-organized criticality: which ones best describe brain activity ?| old_uid | 5473 |
|---|
| title | Stochastic dynamics, chaos or self-organized criticality: which ones best describe brain activity ? |
|---|
| start_date | 2008/10/24 |
|---|
| schedule | 10h |
|---|
| online | no |
|---|
| summary | In awake and attentive animals, the activity of cerebral cortex is highly complex, both at the microscopic (neuronal) or macroscopic (population) level, but it is not clear what type of dynamics is associated with this activity. We review here experimental and theoretical/computational evidence for three types of dynamics in brain activity: stochastic, chaotic and self-organized critical (SOC) states. Global variables, such as the EEG or LFPs, show evidence for chaotic dynamics, but they also display 1/f frequency-scaling, which is more consistent with SOC states. However, the analysis of interspike intervals (ISI) fails to reveal the power-law scaling associated with SOC, but rather displays exponential scaling, more characteristic of stochastic processes. These contrasting observations can tentatively be resolved using modeling, and in particular the "asynchronous irregular" states of activity in randomly-connected networks of integrate-and-fire neurons. Such network states are associated with high-dimensional chaos, but they can also display apparent stochastic dynamics in single neuron activity, and exponentially-distributed ISIs. The 1/f scaling of LFPs can be explained by the frequency-filtering properties of the extracellular electric signals. Taken together, these results are consistent with the idea that brain dynamics stem from high-dimensional chaotic states. |
|---|
| responsibles | Bourgine |
|---|
| |
|