The 35-year-old was diagnosed with progressive, degenerative disease called spinocerebellar degeneration 13 years ago and is paralyzed from the neck down, and cannot voluntarily move her arms or legs — a condition called tetraplegia.
She was surgically implanted with microlectrodes that recorded brain waves in the motor cortex, which a computer program translated into movement commands.
In the lab, she consistently improved on a series of reaching and grasping tasks over 13 weeks using the robotic hand.
"The participant did the manoeuvres with co-ordination, skill and speed almost similar to that of an able-bodied person," the study's lead author, Prof. Andrew Schwartz of the University of Pittsburgh and his co-authors said in Sunday’s online issue of the medical journal The Lancet.
The computer algorithm mimics the way the unimpaired brain controls limb movement, the researchers said.
Over time, she was able to do the tasks, such as grasping and moving items of various shapes and sizes, with a success rate of up to 91.6 per cent.
"The main advantage of the prosthetic is that it looks like a human arm, and has much of the same capability as a human arm in terms of flexibility of movement at multiple joints, including in the hand," study co-author Michael Boninger, chair of the university’s department of physical medicine and rehabilitation, said in an email.
Boninger said the main achievement was the high level of control in the brain/computer interface.
The researchers hope their work will help individuals with tetraplegia or upper limb amputation to regain natural behaviours to interact with the world around them.
Woman focused on goal, not process, to move objects
The experiment was the first time a person performed the activities better than a non-human primate, Prof. Grégoire Courtine of the Swiss Federal Institute of Technology in Lausanne said in a journal commentary.
The control mechanism is remarkably similar to how the central nervous system is organized in mammals, Courtine noted.
The woman said she only focused on the goal of the action, not the details of how to move the object.
"This bioinspired brain–machine interface is a remarkable technological and biomedical achievement," Courtine wrote.
But there are challenges for the technology's use outside the lab. In the experiment, highly skilled engineers tuned the decoding algorithm.
The technology also still needs a safe, low-power and wireless system to record brain activity properly for decades, Courtine said.
The research was funded by the U.S. Defence Advanced Research Projects Agency, U.S. National Institutes of Health, Department of Veterans Affairs and UPMC Rehabilitation Institute.