You are here: Home News Publications Faraway, so close!
Document Actions

Faraway, so close!

Combination of experiment and computation sheds light on the role of long-range connections in the brain

Faraway, so close!

Image courtesy: www.kaviarasu.com

The brain consists of large, interconnected networks of neurons. As a consequence, a single neuron can have connections to many thousands of other neurons. This means that it can receive input from many thousands of neurons and, likewise, can provide output to many thousands of neurons, not necessarily the same ones. Since the connectivity structure of these networks crucially determines how the brain processes information, scientists search for more detailed information about this structure as well as for more abstract models of its function. Until recently, it was assumed that neurons in the cortex are arranged in a columnar architecture, such that most of the incoming-connections of a neuron stem from closely neighboring neurons, either within the column the receiving neuron lives in, or from neurons in adjacent columns. Long-range connections, scientists thought, make up only a small part of the overall connectivity structure and therefore, play only a minor role in the processing of information in the brain.

In a recent study published in the journal ”Cerebral Cortex”, researchers from the Bernstein Center Freiburg and Cluster of Excellence BrainLinks-BrainTools used a combination of experimental (physiological, anatomical) and computational techniques to investigate the role of long-range connections in the primary somatosensory cortex of rats – a brain area involved in processing information about bodily sensations. The researchers measured the spatial extent of the connectivity and the physiological properties of long-range, horizontal connections to one of the major classes of neurons in the neocortex. These data were then incorporated into a computational model simulation, which showed that these long-range connections could indeed improve the information processing capabilities of such networks by reducing “noise correlations” (comparable to a radio’s bad reception), and by improving the signal detection (i.e. the sensory input, or, to keep the metaphor, the signal of a radio channel) in the network.

Philipp Schnepel and colleagues could thus show that the number of long-range connections to an important class of neurons in the cortex has until now been drastically underestimated and that these long-range connections can play a much more prevalent role in information processing than previously thought. Furthermore, the abundance of such long-range connections has important implications for external stimulation of these brain regions for therapeutic and neurotechnological applications. The authors study such applications in the context of the Cluster of Excellence BrainLinks-BrainTools: On the one hand these new data imply that the effect of a therapeutic stimulation of the brain could be more reliable than expected, while on the other hand the responses will be more widespread and, therefore, less specific.


Original publication:

Philipp Schnepel, Arvind Kumar, Mihael Zohar, Ad Aertsen and Clemens Boucsein (2014) Physiology and Impact of Horizontal Connections in Rat Neocortex. Cerebral Cortex, doi: 10.1093/cercor/bhu265


Personal tools