At the Society for Neuroscience meeting in San Diego last week, a group of researchers presented data on a speech prosthesis which they say could soon enable a paralyzed man to talk again.
The device, which consists of 3 gold recording wires, was implanted into the brain of Eric Ramsey, who was completely paralyzed in a car accident 8 years ago. Ramsey is said to be "locked in" - he is fully conscious but is unable to communicate in any way.
The implant is located 6mm below the surface of Ramsey's brain, in a part of the left premotor cortex, where it records the activity of 41 cells involved in speech production. Lead wires connect the electrodes to an amplifier and FM transmitter attached to the outside of the skull. The recordings are sent wirelessly to a computer.
The device is what the researchers call a neurotrophic electrode. The electrodes are attached to a glass cone which is filled with growth factors, and which therefore encourages axons and dendrites to grow onto it. This integration with the surrounding tissue makes the implant stable. It has, in fact, been embedded in Ramsey's brain for the last three years.
For this time, Ramsey has been imagining saying simple vowel sounds, such as "ee" and "oo". By analyzing the recordings taken over such long periods, the researchers believe they have determined the brain activity correlated with the planning of each sound.
Currently, they say they can accurately determine which sound he is thinking of about 80% of the time, and have started feeding the data into speech software which translates it into sounds. Eventually, when the range of sounds that can be produced is expanded, Ramsey and others like him could use the device to hold real conversations.
Related:
Brain & Behavior
Medicine & Health
Comments (2)
These cyborg implants are reverse engineered to map out the brain's activity. Do they have to be reverse engineered for each person, does everybody have different brain activity for the same sounds, or does the software simply have to be 'calibrated'? I don't doubt that they will succeed in making this produce basic sounds needed for coherent speech, scientists were were successful in translating the brain's arm movement activity to moving a mouse pointer. Brain-computer interface technology is so cool.
Posted by: sam lidester | November 15, 2007 11:42 PM