guia do novo coronavirus
Imprimir Republish


Banana connection

An electric signal sent directly to the brain gives monkeys a clue about where to find food

Duke UniversityAn experiment with Northern night monkeys has proved that they are capable of interpreting multiple electrical signals transmitted directly to their brains, thereby improving on the results achieved in a guessing test. Conducted by a team working with Brazilian neuroscientist Miguel Nicolelis, this study is an important step towards understanding how the brain works and in the development of technology that will make it possible to  build easily controllable robotic prostheses which can reproduce some characteristics of the human body, such as touch or sensitivity to temperature. The researchers bet that if it went well with monkeys, it should work with humans, as the brains of both have similar structures and work in similar ways.

“There’s great interest in developing this area because it’s believed that micro-electrodes may serve as an artificial channel to transmit sensations to the brain that have been lost through neurological damage or to transmit feelings from a prosthetic limb”, says Nicolelis, who years ago was going in the opposite direction: he used the electrical signals captured by micro-electrodes implanted in the brains of monkeys to move a robotic arm.

This time, in the Duke University lab in the USA, the researchers submitted two female Northern night monkeys (Aotus trivirgatus) to 40 sessions of a test that at first seemed like a child’s game, but gradually became complex and sophisticated. In the first stage Nicolelis and neurophysiologists Nathan Fitzsimmons, Weying Drake and Mikhail Lebedev simply trained the monkeys to point to the back of two wooden doors where a piece of banana was hidden. But they gave them a clue beforehand: they let them see the door where the food was  hidden through a glass barrier.

When they became expert at finding the snack, the researchers began to make things more difficult. Instead of opening the door to the box containing the banana so they could see it, they started activating a small vibrator on the shoulder corresponding to the door with the food. If the monkeys got it wrong a person from the team immediately removed the banana from the other door to avoid their using this failure as a clue to where the food was.

The following step was to substitute the small tremor, which worked like a light touch on the shoulder, with electrical signals transmitted directly to the region in the brain – the somatosensory cortex – that interprets the sensations of pain, cold and heat received from the hand. During the transition period Nicolelis and his team repeated the experiments, supplying the two types of clues simultaneously – the vibration on the shoulder and the electrical signals to the brain – until the monkeys learned to associate them. At the test session 35 of the two volunteers with large brown eyes were getting it right more than 85% of the time, showing that they had learned to use the clue provided by the researchers.

As the electrodes were only implanted in the left side of the brain, which is responsible for the sensitivity of the right hand side of the body, Nicolelis resorted to a strategy that was somewhat more refined. With the help of a computer he started sending two electrical pulse sequences to the electrodes in the brains of the monkeys, each of which lasted four seconds: either short or long pulses. The sequence of short pulses, which lasted 150 milliseconds, interspersed with 100 millisecond intervals, indicated that the correct door was the one on the right. The sequence of long pulses, on the other hand, (lasting 300 milliseconds, with intervals of 200 milliseconds) meant that the food was behind the door on the left. This time the monkeys achieved a good performance level more quickly (by the eighth session), as the researchers describe in an article published in the Journal of Neuroscience, May 23rd issue.

After they had learned to differentiate the codified clues in the form of long and short pulses – information which the researchers called temporal – the monkeys had to deal with yet another challenge: to identify the order in which the four micro-electrodes were being activated. At this stage of the experiment, instead of sending the pulses simultaneously to all electrodes, Nicolelis started to activate them in sequence – A, B, C and D, meaning “right door”, and D, C, B and A, meaning “check the left door” -, thus associating a spatial characteristic with the information; the monkeys managed to interpret this even faster.

“We believe that this species of monkey can understand at least ten forms of spatial and temporal stimuli, a number that humans could undoubtedly  exceed”, says Nicolelis, who is still doing research at the Polytechnic University of Lausanne, in Switzerland, and at the Natal Edmondo and Lily Safra International Institute of Neuroscience in Rio Grande do Norte.

These results are important for us in terms of understanding how the brain responds to these stimuli and uses them to control behavior. According to Nicolelis, although it is almost impossible to reproduce the natural functioning of neurons by means of electrical pulses sent via electrodes implanted in the brain, the accuracy of this information transmitted may be sufficient to improve the quality of life of those who use robotic prostheses. The other good news is that the microelectrodes do not damage the nervous system, even when used for long periods of time.