One beauty in the design of BCI projects is the “learning” part of the project – the process of machine learning is one of the most compelling discussion topics we see in this field. How do we get the computer and the brain to operate synergistically? While you might not succeed at teaching a machine how to do your holiday shopping this year, here’s some interesting BCI news we received lately from the UK.
1/ GeekCon 2011: What can you do with an Emotiv EPOC headset?
Some highlights from the seventh annual GeekCon, the exclusive creative gathering run by three Israeli technology entrepreneurs — Ilan Graicer, Eden Shochat, and Nimrod Lehavi – were published in the December 2011 issue of Wired Magazine (UK) this week. Nimrod (pictured), a software developer, played around with the Emotiv EPOC headset.
What he found: After writing a program to capture Emotiv’s outputs as keystrokes on a laptop and transmitting them as commands connected to a remote and toy car, the circuitry could handle simple directions (forward, back, left, right). The system had trouble distinguishing particular emotions, but an “I hate you” - caused the car to speed away. Lesson? Road rage might help out with hands-free commuting.
Other tech projects that explored out-of-the-box thinking included: robots that shoot tomatoes, and a beer cooler you play like an instrument. See for yourself here.
2/ Simulation-based model of a limb on motor imagery BCI reveals insight into BCI use for paralysed individuals
In the social discussions that take place on the web around this field, we’re always listening to what bright curious minds are getting up to. Recently, a student from The University of the West of England shared with us his EEG based BCI research out of Bristol. And sharing is caring!
It is understood that similar brain areas are activated when a physical movement is watched, as when it is being made by the same individual. The firing of these “mirror neurons” (used in Ramachandran’s Mirror Neuron Therapy on paralysed individuals), inspired a novel pilot study at UWE. Using the BCI2VR system developed by Dr. Ou Bi, the team tested the effect of a simulated limb on motor imagery-based BCI. Participants were provided with an identical moving arm rather than their own, and it created a faster and more accurate EEG model of their motor imagery.
The application? “If the results of the experiment are valid, then we could have a very simple and inexpensive way of training paralysed individuals to use a BCI”. Simon Oxenham contributes to the Neurobonkers blog, and the full research report is available here.
To wrap up the news, keep an eye out for InteraXon at Macau IT Week!
