Researchers from University of Minnesota have taken the brain-computer interface a bit further by giving the user full control of the movements of a helicopter in a virtual environment. 
Although experiments with brain-computer interfaces have been going on for quite some years already, most tasks have involved one dimensional or two dimensional movement (we recently covered mind controlled slot car racing). Three-dimensional control has been achieved with the use of invasive brain electrodes, but non-invasively this has been notoriously hard to achieve.
Here the researchers have achieved just that by using the EEG signal from a 64-channel EEG cap to record and decode sensorimotor rhythms induced from motor imaginations. The subjects first went through a training utilizing left/right arm, leg, tongue, and rest imaginations to move a cursor first in 1D and the n in 2D, before moving on to the 3D helicopter task. They even managed to differentiate between gross and fine movements on a continuous scale, allowing for fast and precise control of helicopter movement.
As a final task, the subjects controlled the helicopter with the goal of flying through randomly positioned rings. Impressively, they successfully acquired over 85% of presented targets. Although the authors foresee applications in neuroprosthetics and rehabilitative medicine, we are envisioning controlling real-life model helicopters or drones instead. The results were published online in PLoS ONE.