Exoskeleton hand lets quadriplegics manipulate objects with power of their mind (VIDEO)
German scientists have completed the early trial of a robotic hand that allows people with severe spinal cord injuries to write their signature with a pen, grasp a fork to eat food, or use a credit card, with the help of nothing more than a signal sent from the brain.
“The patients were amazed by what they could do with the system,” Surjo Soekadar from the University Hospital of Tübingen, who led the experiment, told New Scientist. “Previously, they couldn’t have a meal with a knife and fork, so changing that was amazing for them.”
An elastic cap, similar to those worn by swimmers, but fitted with electrodes, was put on the heads of the six participants in the trial. Whenever they wanted to clasp something with their hand, an action they were otherwise unable to perform, their brain produced a recognizable electric pattern, which was detected by the cap, and sent to a control box.
The box passed the signal onto a set of actuators, and finally the exoskeleton, a Terminator-like contraption fitted onto the fingers, capable of executing precise and smooth movements.
It took all participants between eight and 10 minutes to train themselves to use the device comfortably. The system was also robust enough to function in public places – where there were distractions that could distort the brain signal if the control box misread it.
Separately, all of these constituent parts of the mechanism had been tried, but previously it had been impossible to build a practical device from them – particularly one that wouldn’t endanger the users, who have no easy way of coping with physical emergencies.
“Normally, systems like this have around 90 per cent accuracy. That might sound good, but in principle it means that, one time out of ten, a patient could drop a cup of coffee on themselves,” said Soekadar, the co-author of the study, which has been published by Science Robotics, a new academic journal.
To minimize such dangers, Soekadar’s team inserted a delay of 1.5 seconds between the command and the action, and placed another electrode near the corner of the eye. This detector would respond to a rapid sideways eye movement – one people never normally make – which would serve as a virtual ‘cancel’ button if users were about to make a mistake.
Other than the need for a carer to fit the cap, the system could leave quadriplegic users autonomous for hours. In earlier experiments, electrodes had to be physically implanted into the user’s brain to be effective, an approach that could lead to infection, rejection and other dangers linked with such invasive procedures.
“There is almost no risk and a clear benefit,” David Guiraud, neural engineer at the University of Montpellier in France, told the New Scientist, of the innovative exoskeleton.
But Soekadar admits that the system is still not as accurate as an implant, and cannot, for example, manipulate individual fingers. He plans to add context recognition mechanisms that would better help the exoskeleton identify where it is – in a public or private space, for instance – and what sort of action needs to be performed, detecting what object is in front of the bionic hand.
“In the next couple of years, systems like this will become commercially available. But it will take a few more years before they can become really smart,” said the researcher.