Currently available neuromotor assistive technologies are limited by low performance. They are typically slow and inaccurate, and can require the user to adapt the neuronal activity in their brain, which requires months of intensive training. However, the work of Hochberg, Santhanam and their respective colleagues suggests that there are solutions to these problems.
Hochberg and co-workers developed a brain–computer interface (BCI) that used a 96-microelectrode array implant to measure neuronal firing patterns directly in the motor cortex. The clinical usefulness of the team's BCI hinged on whether movement signals persist in the motor cortex long after their communication route to the brain has been disrupted or severed. The participant in the trial was a tetraplegic man (M.N.) who had sustained an injury to the C3 and C4 vertebrae 3 years before he took part in the trial. Amazingly, neuronal firing patterns in M.N.'s motor cortex were similar to those generated in intact monkeys. To correlate these existing neuronal firing patterns with a specific movement, the authors coupled the implant to a computer and recorded activity while M.N. imagined making a range of movements with his hands and arms. By analysing movement-dependent variations in neuronal ensemble activity, the authors decoded the firing patterns, which enabled M.N. through thought alone to control a 'neuronal cursor' to open an e-mail, play video games and manipulate an object using a prosthetic hand, while simultaneously holding a conversation!
This is a preview of subscription content, access via your institution
