Lamprey cyborg sees the light and responds.

The $6-million eel it ain't. But researchers who have taken the unprecedented step of connecting a brain, in this case a sea lamprey's brain, to a small mobile robot say they've got a roving fishbot that may someday lead to better prosthetic devices for humans.

In the meantime, the two-wheeled robot that scoots about the lab bench in response to light may help scientists better understand how an animal's nervous system and a machine can communicate with each other.

Led by Sandro Mussa-Ivaldi, a computational neurobiologist at Northwestern University Medical School in Chicago, the researchers described their findings this week at the 30th annual meeting of the Society for Neuroscience in New Orleans.

"I'm very impressed with this work," says Stephen P. DeWeerth, a biomedical engineer at Georgia Institute of Technology in Atlanta. "It combines the realism of biology with the controllability of electronics."

To create the cyborg, Mussa-Ivaldi and his team removed a lamprey's brain and part of its spinal cord and placed them in a refrigerated, oxygen-rich saline solution. Then, they rigged small wires to bring electrical signals from optical sensors mounted on the robot to the disembodied brain's vestibular system, which normally enables the animal to distinguish up from down.

When the robot's sensors detected light, the lamprey's brain interpreted the signals they sent as conveying a certain orientation in the water, the scientists speculate. Electrical impulses that normally would have moved along nerves to the animal's muscles instead traveled along a second set of wires to the robot's wheels.

Depending on the electrodes' placement in the brain tissue, the signals caused the robot to wheel toward or away from the light or to travel in a circle or a spiral. In experiments with more than 40 lamprey brains, approaching the light was the most common response.

Mussa-Ivaldi says the disembodied brain's continuous response to the ever-changing signals from the robot's optical sensors is the first example of two-way interaction between neural tissue and a robot. Studying the behavior of this simple fish-robot hybrid may help researchers develop more complicated systems that combine biological and electronic elements, he notes.

"Scientists don't have a clear understanding of how to develop a two-way communication between a brain and an artificial limb," Mussa-Ivaldi told SCIENCE NEWS. "We need to understand how the nervous system and a machine may talk to one another."

The researchers chose a lamprey's brain to control their robot for several reasons, says Mussa-Ivaldi. For one thing, the neurons are large and easily identified. Also, the tissue can be kept alive outside the animal's body in a cold saline solution for weeks. Moreover, he notes, the lamprey's central nervous system can compensate for a lesion in the vestibular system. When some of these cells are damaged, connections between the neurons somehow rearrange so that the lamprey still swims normally.

In their next set of experiments, the researchers plan to examine the brain's flexibility after such rewiring. They'll simulate a lesion in the lamprey's nervous system by electronically blocking the input to the brain tissue from one of the robot's sensors. After the brain has adjusted, the scientists will restore the signal and observe the robot's behavior. Such a before-and-after comparison would be impossible with a surgically induced lesion but easy to obtain via reversible electronic simulation, Mussa-Ivaldi says.

Any organism that receives an implant or a prosthesis connected directly to its nervous system will have to adapt to the device's presence and learn to communicate with it, says DeWeerth. Characterizing the interactions between simple biological and electronic materials is a fundamental step in building true hybrid systems, he notes.