Scientists from Switzerland with colleagues from other countries took a step used to treat a range of spinal injuries where patients lose mobility in their limbs. The researchers managed to bypass damaged nerve tissue in the spinal cord by creating a wireless digital bridge between the brain and spinal cord underneath the damaged area. But even without the help of machine learning, it was necessary to be able to recognize movement thoughts.
The project was led by researchers from the Swiss Federal Polytechnic School in Lausanne (EPFL). A 38-year-old man who suffered a cervical spine injury in a fall from a bicycle ten years ago was paid. Previously, he participated in a program to support the rehabilitation of people with spinal injuries. In particular, he used the method of epidural stimulation of the spinal cord, in which an implant with electrodes is placed in the spine and a stimulator is sewn under the skin. Such a platform, based on the readings of the movement sensors in the stimulator, creates impulses in the responsible areas of the spinal cord, forcing the muscles of the limbs to work, but the ability of a person to move is very, very limited.
Because the patient still had electrodes in his spine (on the spinal cord), the scientists decided to give them a control signal from the brain. To do this, it was necessary to organize a digital radio bridge, since the nerve tissue between the spinal cord and the brain was torn as a result of trauma. In order to read signals from the brain, sensors with their arrays of electrodes were implanted in the patient’s skull. The electrode control unit received external inductive radio current with a frequency of 13.56 MHz, and the brain activity read was transmitted by another antenna – a decimeter antenna with a frequency of 405 MHz.
The data was received and decrypted by a receiving device (perhaps a laptop) that the patient had to carry in a backpack on his back. First, the algorithm was trained to recognize brain activity in response to commands to perform specific leg movements, and then it was trained to synchronize the patient’s desires to move the limbs with signals sent to the spinal cord and onward the target muscles are routed to the legs.
Through training, the digital interface helped the patient do what he was unable to do after the injury — walk over rough terrain and balance with crutches. The platform also worked well at home, and not just under medical supervision. In addition, part of the neural pathways in the brain were able to rebuild and the patient was able to perform a number of actions without artificial stimulation. One day the researchers find out in their Article V Nature, such technologies will be able to return people with spinal injuries to an active life. If it works for one patient, it can be repeated for others.