The scientists created a unique brain prosthesis. It operates on a plug and play basis. It was first demonstrated in a paralyzed person. The brain prosthesis was created by the specialists from the Institute of Neurology at the University of California. They developed a brain-controlled prosthetic limb and demonstrated a machine learning techniques that allowed a paralyzed person to control a computer cursor using only their own brain activity.
In recent years, a great progress in that area has been made. But all existing systems had to be overloaded and reconfigured almost every day. They couldn't use the brain's natural learning processes. It was like learning to ride a bike starting from scratching every day.
An adaptation of an artificial uninterrupted machine learning with complex schemes of long-term actions, that experience had not exist in science before. Karunesh Ganguly, MD, believes that the achievement of a plug and play complex shows the value of the set of electrodes used in the new design.
Its matrix includes a very small electrode array. It is surgically injected onto the surface of the brain. The electrodes guarantee long-term stable recording of neural activity. The scientists approved the system as a monitoring method for studying seizures in patients with an epilepsy.
In clinical trials, the new device was installed in a man with four-limb paralysis. He was able to use an embedded implant to control a cursor on a computer monitor. The scientists developed a special algorithm that aims to use machine learning to match the brain activity recorded by the electrodes with the desired movements of the user's cursor.
The lessons were long and not easy, as it seemed from the first time. But the algorithm was refined and received a constant interaction between brain signals and increased performance for many days.