Just in time for the influx of sedentary Oprah viewers, [Adam Wilson] built a brain interface that allows you to post Twitter messages. The electrode cap monitors the user’s brain functions to determine where they’re looking. The display slowly flashes each letter in the alphabet. The user focuses on the letter they want and when it flashes the cap can pick up the resulting impulse. It’s a long process and the average user can only do ten characters a minute i.e. 14 minutes to use all 140 characters in a Twitter post. It’s interesting research and shows how far we still need to go with neural interfaces. The researchers note that Twitter’s forced brevity levels the playing field between locked-in patients and normal users. A video of the device in use is available on the NITRO blog.
Related: KanEye tracking system
Yesterday we linked to an OCZ Neural Acutator Interface teardown. Several in the comments wanted to know more about the sensor electrodes. Check out the OpenEEG project and OpenEEG mailing list for information on sensing, amplifying, and recording brain activity (EEG). The OpenEEG project maintains an open source Simple ModularEEG design. Two other open source variants of the ModularEEG are the MonolithEEG and [Joshua Wojnas’] Programmable Chip EEG BCI. All three projects use Atmel microcontrollers, with designs in Cadsoft Eagle.
Brain activity is measured using passive or active electrodes. Passive electrodes require a conductive paste to make proper contact with the skin (examples: 1, 2). Active EEG sensors don’t need conductive goop because they have an amplifier directly on the electrode (examples: 1, 2, 3).
[via anonymous reader, comments]