Neuroscience: Piloting a virtual quadcopter using a brain–computer interface

A surgically implanted brain–computer interface can detect and decode finger movements in a person with paralysis, allowing them to play a video game, reports a study in Nature Medicine.

More than 5 million people in the US live with severe motor impairments. Although many of the basic needs of people with paralysis are being met, there remain unmet needs for social and leisure activities, such as video games. Brain–computer interfaces have been recognised as a potential solution for motor restoration, but current examples of this technology have struggled with complex movement such as individual finger movements, which could help with activities like typing, playing musical instruments, or using a video game controller.

Mathew Willsey and colleagues developed a brain–computer interface capable of continuously recording the electrical activity patterns of multiple neurons in the brain to translate complex movements. The interface was implanted in the left precentral gyrus — the brain region responsible for hand movement control — of a person with upper and lower extremity paralysis. Neuronal activity was recorded as the participant observed a virtual hand performing various movements, after which the researchers used machine learning algorithms to identify the signals linked to specific finger movements. Using these signals, the system was able to accurately predict finger movements, enabling the participant to control three highly distinct finger groups, that included two-dimensional thumb movements, in a virtual hand. This system achieved a level of movement precision and freedom greater than previously possible.

The authors then extended the application of this finger control to a video game. Finger movements decoded by the interface were programmed to control the speed and direction of a virtual quadcopter, allowing the participant to pilot the device through multiple obstacle courses as a part of a video game.