back to the Eurescom home page

 

mess@ge home

 Table of contents
of the current issue
 

Selected Highlights
Simplifying complexity
Developing
usable mobile
phones

 
The Berlin Brain-
Computer
Interface

 
Interview with
Dr  Nico Pals and
Joke Kort from
TNO
 
Speech
technology and usability

 

The Berlin Brain-Computer Interface

 

Prof. Dr. Klaus-Robert Müller
Fraunhofer Institut FIRST
Klaus-Robert.Mueller@first.fraunhofer.de

”If you could read my mind”, Gordon Lightfoot had sung in the 1970s. Three decades later, computers are coming close to doing the job. Numerous research groups in Europe and America are working on concepts for brain-computer interfaces. These “mind-readers” include an interdisciplinary research team in Berlin. By analyzing neural signals, computer scientists from the Fraunhofer Institute for Computer Architecture and Software Technology FIRST and neurologists from the Benjamin Franklin university clinic can determine before the actual movement takes place, whether a person intends to move his/her right or left hand, for example.

The electrical activity in the brain is measured by 128 electrodes affixed to the person’s scalp, as for an electroencephalogram (EEG). The Berlin Brain-Computer Interface (BBCI) volunteers have no need of lengthy training sessions to learn how to control their mental processes. Where most groups require 100-300 hours of subject training, the BBCI requires only 20 minutes of subject training and the computer then learns to correctly interpret the neurophysiological signals within 1-2 minutes given this training data. In fact, the Fraunhofer experts have developed a software programme capable of picking out specific signals among the nebulous mass of information. The computer’s self-learning capacity allows it to identify individual brain patterns and constantly improve its performance.

The BBCI offers a new channel to interact with computers in real-time. There are many possible applications for this human-machine interaction technology. For example, a sort of “mental typewriter” that translates thoughts into cursor movements on a computer screen, allowing paralyzed patients to write texts. The same technique may one day also enable them to control a prosthetic device. Brain-computer interfaces could also spread to the entertainment industry, creating a whole new class of video games. Or they could be integrated in active car safety systems, for instance braking the vehicle in response to the driver’s thoughts.

Please send us your comments on this article.