The Potential of Potentials

Brendan Allison, Ph.D. gave a presentation to a Microsoft Research audience in June 2006, titled Brain Computer Interface Systems: Progress and Opportunities, covering some exciting R&D.

Brain computer interfaces (BCIs) use neuroimaging systems (most commonly EEG) to measure brain activity and process signals to control software. Research is underway on implants/neural prosthetics, but EEG is non-invasive and getting more accessible. The familiar bulky array of 8 to 256 electrodes with conductive gel stuck to a user's head, impractical for portable consumer use, is changing as companies develop caps, glasses and similar small devices that allow freedom of movement and no professional assistance to hook up.

For people with locked-in syndrome from ALS, who cannot move any muscles in their body but remain conscious and alert, BCIs can be their sole means of communication with the world. Thought Translation Devices (it's not quite thought, rather ERPs) are used to write, and patients can even browse the web with Neural Internet. Researchers have created a robot controlled by BCI to function as an assistant for the disabled, and no doubt there are other human-liberating projects in development and refinement.

Games tend to get awed attention (see this clip on YouTube of Pong). There's Mindball, a competitive ball game, a video game purports to treat ADHD, and other neurofeedback games make various claims.

In the arts they've been used for performance drawing, artifact imaging, there's a proposal to create live video, and other nifty creative works.

Then there's the controversial Brain Fingerprinting lie detection method, and other applications too numerous to list in just one post.

Back in the Microsoft Research lecture video, Brendan Allison says electrode arrays cost as little as $1,000 readymade, and can be used with an ordinary computer (even a PDA). Mix in some DIY creativity and who knows what may result? If Make magazine/blog meets the OpenEEG wiki, it's really unpredictable. The future seems wide open.

EEG does have limitations. It takes considerable training to operate a BCI with a biofeedback-type method in which the user relaxes into different states; or to set up an interface based on recognizing objects. The process and interface can be cumbersome.

Considering electroencephalography has been around since the early 20th century, and growing more sophisticated, it seems we'll continue to invent applictions. Allison outlines much of the progress in his talk, so do watch his presentation.

Note that Microsoft has considerately limited viewing of the video to users with Explorer and Windows Media Player. How unfuturistic of them.


Journal refs:

Review Brain-computer interfaces--the key for the conscious brain locked into a paralyzed body. K├╝bler and Neumann,
Progress in Brain Research 2005.:513-25.

Neural Internet: Web Surfing with Brain Potentials for the Completely Paralyzed, Karim et al., Neurorehabilitation and Neural Repair, 2006, 508-515 (free PDF)


Blogger Brendan said...


One of the fun things about BCI research is that it changes so quickly. Since my Microsoft talk last year, a few groups have described commercially available dry electrode systems for much less than $1000. I have not yet seen or tested any of them, but have talked to credible researchers at conferences.
--Brendan Allison

28/8/07 03:37  

Post a Comment

<< Home

eXTReMe Tracker