The Technology Review article Better Brain-Wave Analysis looks at start-up ElMindA that is trying to find new quantitative methods for broadening the clinical use of EEG.
The company has developed a novel system that calculates a number of different parameters from EEG data, such as the frequency and amplitude of electrical activity in particular brain areas, the origin of specific signals, and the synchronicity in activity in two different brain areas as patients perform specific tests on a computer.
This description doesn't sound very novel, but I've always felt that EEG analysis has tremendous clinical potential. This is particularly true for rehabilitation purposes (like the stroke example) and EEG-based communications for paralysis patients.
I am skeptical of "objective diagnosis" claims for things like attention-deficit hyperactivity disorder (ADHD) though. In the 1980's EEG Topography was thought to be able to distinguish some psychiatric disorders. These claims were never proven to be true.
I'm not saying that new quantitative techniques like those being developed by ElMindA are comparible to the old EEG "brain mapping", but significant clinical validation will be required before they can be used clinically.
UPDATE (6/3/09): Another related article (also from Technology Review) Reading the Surface of the Brain (and cool picture):
UPDATE (6/10/09): The BrainGate project at Massachusetts General Hospital has recently started clinical trials that may help paralyzed patients. The on-going MGH project:
... the ultimate goals of which include "turning thought into action": developing point and click capabilities on a computer screen, controlling a prosthetic limb and a robotic arm, controlling functional electrical stimulation (FES) of nerves disconnected from the brain due to paralysis, and further expanding the neuroscience underlying the field of intracortical neurotechnology.
UPDATE (7/5/09): This is more related to Brain Control Headsets, but if you're interested in developing your own EEG-based controller you should check out An SDK for your brain. The free NeuroSky MindSet Development Tools along with a $200 headset will get you started developing your own "mind-controlled" game. Good luck with that!
Hello Bob,
I read the Technology article before, and had similar thoughts. At least, I feel it will take some time before we can see practical applications. I am sure there will be some type of “interpreting analysis” but it will take a lot more development and as you put it “significant clinical validation”.
It is good to see someone taking the steps to achieve this out there.
Srihari