16.11.2012 Views

Brain–Computer Interfaces - Index of

Brain–Computer Interfaces - Index of

Brain–Computer Interfaces - Index of

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

216 D.M. Taylor and M.E. Stetner<br />

eventual human testing <strong>of</strong> intracortically-controlled whole arm and hand prosthetics<br />

within the next few years.<br />

Human testing <strong>of</strong> intracortical BCIs has already been going on since 1998.<br />

Dr. Phil Kennedy and colleagues received the first approval from the US Food<br />

and Drug Administration (FDA) to implant intracortical electrodes in individuals<br />

with “locked-in syndrome” who are unable to move. These early implant systems<br />

recorded only a handful <strong>of</strong> neurons, but locked-in individuals were able to use the<br />

firing patterns <strong>of</strong> these neurons to select letters in a typing program [18].<br />

In the Spring <strong>of</strong> 2004, Cyberkinetics Inc. received FDA approval to test 96channel<br />

intracortical microelectrode arrays in people with high-level spinal cord<br />

injuries. They later also received approval to evaluate this intracortical technology<br />

in people with locked-in syndrome. The results <strong>of</strong> their initial testing showed that<br />

the firing activity <strong>of</strong> motor cortex neurons was still modulated with thoughts <strong>of</strong><br />

movement even years after a spinal cord injury. These researchers also showed that<br />

the study participants could effectively use these intracortical signals to control a<br />

computer mouse, which the person would then use for various tasks such as sending<br />

emails or controlling their TV [19]. The study participants were easily able to control<br />

the mouse without much concentration, thus enabling them to talk freely and<br />

move their head around while using their brain-controlled computer mouse. The US<br />

National Institutes <strong>of</strong> Health and the Department <strong>of</strong> Veteran’s Affairs continue to<br />

strongly fund research efforts to develop intracortical BCI technologies for severely<br />

paralyzed individuals.<br />

7 What Else Can We Get from Intracortical Microelectrodes?<br />

So far, the studies described in this chapter have focused on decoding information<br />

from the firing rates <strong>of</strong> neurons, which are usually re-calculated each time the braincontrolled<br />

device is updated (e.g. rates calculated by counting the number <strong>of</strong> times<br />

a neuron fires over about a 30–100 millisecond window). However, useful information<br />

may also be encoded in the fine temporal patterns <strong>of</strong> action potentials within<br />

those short time windows, as well as by the synchronous firing between pairs <strong>of</strong><br />

neurons. Studies have shown that these two aspects <strong>of</strong> neural firing are modulated<br />

independently while planning and executing movements [20].<br />

Over the course <strong>of</strong> planning and executing a movement, the timing <strong>of</strong> spikes<br />

across neurons changes back and forth between being synchronized (i.e. spikes from<br />

different neurons occur at the same time more <strong>of</strong>ten than would be expected by<br />

chance) and being independent (i.e. two neurons fire at the same time about as <strong>of</strong>ten<br />

as you would expect by chance). What these changes in synchronous firing represent<br />

or encode is a much debated topic in neuroscience. A study by Oram and colleagues<br />

suggested that the changes in spike timing synchronization do not encode any directional<br />

information that isn’t already encoded in the firing rates [21]. Instead, spikes<br />

become synchronized when a movement is being planned. The level <strong>of</strong> synchrony<br />

is highest at the end <strong>of</strong> movement planning just before the arm starts to move [22].

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!