"Robot avatar" for paraplegics
Switzerland has created a "robot avatar" for completely paralyzed people
Scientists from the Federal Polytechnic School of Lausanne (Switzerland) presented the final result of their long-term project to create a remotely controlled mobile robot, which can be commanded using the "power of thought". A description of all the developed projects is published in a special issue of Proceedings of the IEEE dedicated to brain-computer neural interfaces (Muller-Putz et al., Towards Noninvasive Hybrid Brain–Computer Interfaces: Framework, Practice, Clinical Application, and Beyond).
19 volunteers participated in the experiments and testing of robots – 10 healthy people and 9 completely paralyzed. For several weeks, scientists analyzed their electroencephalogram, event-related brain potentials and electromyogram. All the raw data received was translated into a form understandable to the computer.
Basically, the scientists used sensorimotor Mu-rhythm – arch-shaped waves recorded in a calm state in the areas of the brain responsible for movement and sensory sensations associated with the body. It is known that in the case of activation of the motor cortex, the amplitude of the Mu-rhythm decreases, and even if the action itself is not performed, but only appears. This phenomenon can be tracked and taught by a computer to recognize it, which makes it possible to control something by mentally imagining any movements.
The second indicator used is P300. This is a late component of event-related potentials that occurs, for example, when it is necessary to make a decision when choosing an action in a complex sensorimotor reaction (Suppose we have one blue flash for ten red ones, with different probabilities. It is necessary to press the button only when it is blue – in this case, in response to each flash, the evoked potential will be formed, and its late component P300 will display the decision-making of the subject).
Schematic representation of the event-related potential. P3 is a component of P300.
Image: Wikimedia Commons
However, the most innovative was the separation of control between a person and a computer. The robot uses its own sensors to bypass and overcome obstacles, giving a person the opportunity to make "general" decisions, such as turning in one direction or another, entering a room, etc.
After analyzing the EEG, SSP and EMG, with the application of machine learning algorithms to them, about 10 more days of user training were required. After that, they could control two modifications of the robot.
The first modification consists of a self-propelled trolley with a monitor and a video camera. The patient can see with the help of the camera where the robot is going, and who he meets. Thanks to the Skype program, everyone can see the patient's face on the monitor and talk to him. The second modification is a self-propelled wheelchair controlled by the "power of thought" in order not to overload the patient, it is also equipped with sensors and sonars that allow the robot to avoid collisions on its own.
Both modifications can potentially bring paralyzed people to a whole new level of independence and self-service. A mobile robot will allow bedridden people to walk and communicate with others without leaving home, and a wheelchair will already move independently.
In 2012, scientists managed to create a robot arm controlled by mental commands, with which a paralyzed patient learned to feed and water herself. In the same year, the French team demonstrated an improved version of ASIMO, which could also be controlled remotely. And back in 1988, Farwell and Donchin used the P300 to create the first "virtual keyboard", which became the most important prototype and herald of all future neural interfaces.
At the moment, the Swiss group is working on transferring its developments to an autonomous, mobile, laboratory-independent mode; creating neurotransplants and exoskeletons that will allow paralyzed people to perform actions with their hands under the control of the brain, but without the participation of muscles.
Portal "Eternal youth" http://vechnayamolodost.ru
26.06.2015