A UK project team hopes to make a key breakthrough in braincomputer interface technology, allowing wheelchairs, robot arms and other devices to be controlled by the human mind more instinctively and effectively.
A research consortium led by Oxford University aims to develop new asynchronous systems that would enable gradual, precise control of external mechanisms, as opposed to the less natural on/off mode of existing synchronous BCI technology.
BCI systems allow the severely disabled to control a cursor or a robotic device simply by thinking about the movement. Most use non-invasive electrodes attached to the user’s head to measure micro-voltage from the activity of the brain’s neurons. These signals are categorised using complex algorithms and are translated into movement.
The new system would ideally use only one electrode. Prof Stephen Roberts of Oxford’s engineering science department, who is leading the project, said the team hopes to develop BCI algorithms sophisticated enough to allow graduated, proportional control of movement rather than a simple on or off switch.
‘It is reasonably simple to use an interface to turn something on or off with a binary control switch. That is well-proven technology,’ he said.
‘But to control a robot arm, that binary control isn’t good enough. You need to be able to control the amount of movement and speed.’
The researchers plan to design a system that works asynchronously — a development that would herald a huge breakthrough, claimed Roberts.
Existing BCIs display a cue on a computer screen at regularly spaced intervals, inviting the user to think left, right, up or down. The brain’s signals are then monitored and the system decides which way the user intends the cursor to move. The asynchronous system will wait passively until the user thinks of a movement and then detect when that change occurs.
According to Roberts this would be a huge step forward. ‘An enormous amount of work is yet to be done to get to that stage,’ he said.
‘To be able to get to an asynchronous system, it will have to be able to both detect the signal and then categorise the movement it is conveying.‘
Detection is itself a difficult process that still needs refining, said Roberts.
‘Normally what you get is a cacophony of noise in the brain of all the neurons firing at once. The aim is to pick up these tiny signals as cleanly as possible without any interference from beneath a mains hum, which is a thousand times bigger than the signals you are trying to get at,’ he said.
The researchers will also develop a ‘one-size-fits-all’ adaptive system, using intelligent technology to quickly adapt to the brain patterns of the user. This would significantly reduce the cost of a unit if the technology is to become commercially viable in the future, claimed Roberts.
The two-year project, which also includes Southampton and Essex universities, has been backed by £180,000 funding from the EPSRC.
Although the immediate benefits of the technology will be for the severely disabled, Roberts believes that it could eventually be used for a number of broader applications as well as in rehabilitative work for people with spinal injuries or strokes.
In the longer term, a future adaptive asynchronous BCI system could have applications in the gaming or entertainment industry or could even be used to control vehicles, claimed Roberts.
Several third parties, ranging from the military to Playstation, have shown interest in the work, but for the time being the researchers’ focus remains firmly fixed on the medical applications of the technology.