The Los Alamos National Laboratory scientists led by physicist Ilya Nemenman said their findings are expected to significantly change theories about the functioning of neural networks, as well as advance development of computers that mimic biological processes.

The researchers, including scientists from Princeton and Indiana universities, used tiny electrodes to monitor electrical impulses generated by motion-sensitive neurons in a fly's brain. The fly was harnessed to a mechanism designed to mimic a fly's natural flight maneuvers. The harnessed fly was also placed in a wooded setting similar to its natural environment.

"In this system, the motion-sensitive neurons emit spikes very often and very precisely," said Nemenman, noting earlier research estimates were derived from measures of electrical impulses from flies that were stationary and in a relatively featureless environment "bordering on the monotonous and predictable."

The research is expected to advance neuromimetic approaches to the development of artificial intelligence, in which computers are designed to mimic biological processes occurring in the brain.

The study appears in the Public Library of Science Computational Biology Journal.

--
Copyright 2008 by United Press International.
All rights reserved.
--