Creating Accurate Mathematical Models of Nerve Cell Activity

October 10, 2007

An international team of scientists claims to have developed a technique that leads to improved mathematical models of nerve cell behavior. Better still, by understanding principles behind the brain's activity, scientists may find ways to improve brain functioning and lessen the chances of a stroke.

Developing better models of nerve cell activity requires detailed information about the phenomena behind cells' generation of electrical signals (or spikes). But because most of the necessary variables are not observable, scientists need a way to flesh them out.

Ivan Tyukin of the University of Leicester and his collaborators developed a method for automatic reconstruction of hidden variables, starting with cells' evoked electrical activity. Cees van Leeuwen, Alexey Semyanov, and Inseon Song (RIKEN Brain Science Institute, Japan) brought expertise in neurophysiological and neuronal activity recordings; and Henk Nijmeijer and Erik Steur (Eindhoven University of Technology, Netherlands) did the work that allowed electromechanical realizations of the models.

"The developed technique will enable the creation of novel brain-machine interfaces," Tyukin said. "The artificial neurons can be easily connected with the machines electronically. On the other hand, being sufficiently close copies of their biological counterparts, they can communicate with the biological cells."

The modeling advance could lead to a better understanding of the principles behind computations in the biological brain. Researchers can use such models to explore alternative ways of manipulating and enhancing brain function.

Source: University of Leicester, Sept,. 21, 2007.