In the future, some researchers hope that people who lose limbs will be able to control robotic prostheses using special high-speed interfaces on the brain computer - as Luke Skywalker did effortlessly in the Star Wars series.
Due to the fact that brain signals are ambiguous and difficult to decode directly, the existing interfaces of brain computers that control robotic limbs do not cope with the assigned tasks and are often too slow, and the movements performed under their control look just awkward.
But all of this may soon change. Last week, a team of doctors and neuroscientists published an article in the journal Nature Medicine about the brain's computer interface, which uses a neural network to decode brain signals and precisely control the movements of a robotic arm.
Decoder for the brain
For the experiment, the researchers obtained consent from a 27-year-old patient suffering from quadriplegia (paralysis of both arms and legs, usually resulting from a spinal cord injury in the region of the cervical vertebrae). A set of microelectrodes was implanted into his brain, and the data taken from the electrodes was fed into various neural networks, which are artificial intelligence systems. AI is generally poorly designed to process brain signals, but it succeeds in finding patterns in large sets of information.
After extensive training for almost two and a half years, neural networks have gotten a pretty good idea of exactly which brain signals are associated with specific muscle commands, and how to transmit mental signals to a robotic limb.
Not only did the neural network allow the patient to move the robotic arm with greater precision and less delay than existing systems, it also worked even better when the researchers allowed it to train. That is, the neural network was able to learn which brain signals corresponded to a particular movement and which hand movements are more effective for a specific task without any prompts from the researchers.
Promotional video:
With a neural network, a volunteer in an experiment was able to pick up and manipulate three small objects with a robotic arm - an ability that is taken for granted in a healthy person, but often not available to those who rely on prostheses to adapt to everyday life.
Serg kite