1x1 Click Here to Go Directly to the Story
Register/Subscribe
Home



FEBRUARY 28, 2001

FROM LE MONDE INTERACTIF

Harnessing Brain Waves to Help the Disabled
A European program called Esprit has technology that can read people's minds, adapt to an individual's thoughts, and turn wishes into commands


By Jean-Philippe PichevinTranslated by Inka Resch
Le Monde Interactif

  STORY TOOLS
Printer-Friendly Version
E-Mail This Story

POLL INSTANT SURVEY >>
With which of the following statements on outsourcing do you most agree?

The benefits of outsourcing to corporate America far outweigh the costs
There's an even split between the drawbacks and rewards
Any benefits are overshadowed by the loss of U.S. jobs
Unsure

VIEW POLL RESULTS >>
  PEOPLE SEARCH

Search for business contacts:

First Name :
Last Name :
Company Name :

PREMIUM SEARCH
Search by job title, geography and build a list of executive contacts

Search by Zoominfo
  Tech White Papers
Most Recent
Most Popular
It seems likely that voice will one day eliminate the need for finger commands on computer keyboards. But as we still marvel at that probability, technology is moving into an even more mind-boggling sphere by allowing commands to be carried out solely through thoughts (see Daily Briefing, Sept. 13, 2000, "Mind-Reading Technology for Hands-Free Computing" and John Williams' Assistive Technology column). And far from being a distant science-fiction fantasy, the prototype of such a program -- designed so that the disabled can command their own electric wheelchairs or write e-mail -- already exists, thanks to scientific research teams in Europe.

The four-year-old program, called Esprit, is run by Italian and Finnish research teams and financed by the European Commission. And it's a unique technological feat. Although other such thought-command technologies exist in the U.S. and Europe, using signals from a brain-wave tracking electroencephalogram, they all seek to create a universal interface -- one that's the same for all users.

INDIVIDUAL INTERFACE.  Esprit, however, led by José del Millán, has chosen to create an individual interface that adapts to each user. "With our system, individuals can choose both the thought category they want in order to generate an action and use their own strategy to make it happen. They can, for example, choose a motor function and imagine that they are lifting their finger," says Millán.

Surprisingly, the system is actually quite simple. Various devices are placed on the head of the user that capture signals from the electroencephalogram. This information is sent to an electronic mechanism that is autonomous and physically independent and that carries out an initial analysis of the information inflow. The data are then amplified and afterwards filtered in order to remove all surrounding electromagnetic noise. Following this, the information is digitized and transferred to the computer program. This is when the action is carried out.

But there are difficulties related to recognizing thoughts in a reliable way, and that is where the software program plays an important role. Information from the captor devices arrives 128 times per second, and the program analyses the data in half-second sequences with the help of algorithms. Thanks largely to a classification process, the system then figures out what category the user's thought belongs to. And if the program finds that it can't discern the thought, it does nothing rather than to carry out a false command. "In addition," says Millán, "our program learns how to decode particular characteristics of the signals sent out by users to better distinguish their thoughts. The machine adapts to users at the same time that users are learning how to utilize the machine."

NEURAL NETWORKS.  Using what are referred to as neural networks, the algorithms work much the same way as the human brain. Based on artificial intelligence technology used today in fields like finance and voice recognition, the concept actually consists of a logical classification process made up of tiny independent units called neurons. These, in turn, are all connected by communication canals, as with the human brain. In the case of the algorithm, each neuron sends messages it has received and calculated from others until it finally achieves thought recognition. And what makes this system special is that it is adaptable, since interconnections between neurons can be modified as the process takes place. In other words, the algorithm learns with experience, and the system actually knows how to adjust to a user's particular patterns.

Today's prototype, which was developed in Italian laboratories, can actually identify five different thought categories. Although the success rate is just 70%, that's offset by a 5% rate of error and the fact that the system makes a decision every half-second. The result: A user's thoughts can be recognized in just 1 to 1.5 seconds. This way, it only takes a few days for users to totally master the system.

Disabled individuals are already testing the system now, which means they will soon be on the road to a much easier and tech-savvy lifestyle. And with the help of programs such as Esprit, the technological divide is progressively closing.


By Jean-Philippe Pichevin
Translated by Inka Resch

Back to Top

[an error occurred while processing this directive]


Copyright 2001, Bloomberg L.P.
Terms of Use | Privacy Policy

Media Kit | Special Sections | MarketPlace | Knowledge Centers
Bloomberg L.P.