Variability in neuronal dynamics

The perceived world is represented in our brains by sequences of spikes. However, repeated presentations of the same stimulus do not result in the same pattern of spikes. This variability is not due to limitations of the neuronal firing machinery. Rather, it is an inherent property of connected circuits that is observed across multiple spatial scales—from the firing of single neurons to activity variations on the level of the entire brain. This variability influences the precision with which brain networks can operate and has major implication on the nature of neural codes. And yet, despite the high presence of variability in brain activity, the sources of this variability and the mechanisms supporting it are unknown.
Why are these sources so challenging to identify? One reason is the highly complex nature of brain connectivity and dynamics. At any given moment, input from multiple brain areas simultaneously impinges on local populations and alters activities therein. Such inputs can be, in principle traced, but practically, due to measurement and analysis difficulties, most variability research has thus far focused on sampling spiking activity from local populations alone.

This project offers a solution to this problem by utilizing the unique ex-vivo preparation in turtles. Using this preparation, we have recently been able to simultaneously monitor the response to sensory input from hundreds of neurons in the cortex, while benefiting from the increased accessibility, stability and high-control conditions typical of in-vitro preparations. The great advantage of this preparation is that it allows us to disconnect all sensory inputs, except vision, thus eliminating uncontrolled sources for variability. Surprisingly, doing so does not eliminate variability suggesting that variability in an inherent property in neural circuits. Our goal is to discover the unknown functions of this variability.