THE race is on to return a natural sense of touch to people who have lost all feeling through paralysis or amputation. One group of researchers has pulled ahead of the competition with the completion of a successful trial, and several other teams are recruiting people to test their technologies.
Recent breakthroughs have allowed people who have lost the use of a limb to move a prosthesis and handle objects. But movement alone isn’t enough. The ability to perceive what you touch is fundamental to precisely controlling and accepting a prosthesis as a part of the body.
That was cracked for the first time last week by Dustin Tyler at the Louis Stokes Veterans Affairs Medical Center in Cleveland, Ohio, and his colleagues. The team has found a way of transmitting long-term, realistic tactile sensations, such as the feel of a cotton wool ball, to two people who lost hands in industrial accidents.
So good is the sensation that one man, on leaving the lab where he tried the prosthetic, said it was like leaving his hand at the door.
In recent years, there have been attempts to recreate a sense of touch by delivering vibrations to a person’s residual limb that equate to pressure on the prosthesis. But this has proved more distracting than helpful and hasn’t been widely adopted. People have also tried attaching electrodes to the inside of residual nerves, but the tingling sensations they produce can diminish over time.
Tyler’s team attempted something more complex. Two years ago, they implanted a cuff of electrodes around the three main nerves in the residual part of the two people’s limbs. These are the nerves that would usually transmit sensory information from the hand to the brain.
Each cuff contained electrodes that could stimulate different parts of the nerves. Wires linked the electrodes to a machine providing a stream of electrical pulses. This was connected to the prostheses the men were already using (see diagram). Eventually, the idea is that the system will operate wirelessly.
“As soon as we stimulated the nerves in the first subject, he immediately said ‘that’s the first time I’ve felt my hand since it was removed’,” says Tyler. As the team switched on each electrode in turn, the men felt a tingling sensation as if it were coming from the tip of their prosthetic thumb, then the tip of the index finger and so on until touch sensations were created across the whole of their prosthetic hands.
With a real hand, touching distinct objects results in different patterns of nerve activity. To mimic this, the team altered the frequencies and intensities of electric pulses. After a healthy dose of trial and error, the first subject said “that’s not tingling any more, it feels real”.
Eventually, the team was able to deduce the pattern of stimulation required to recreate many realistic tactile sensations, such as a pulse, the feeling of pressing a finger very lightly on the tip of a ball-point pen, of someone stroking a finger, and even the sense that a cotton wool ball was being lightly rubbed on the skin.
No more juice
This is only half the battle though. For a prosthesis to be useful outside the lab, it needs to provide a wearer with sensations appropriate for different scenarios. Sensors on the prosthetic limb attempt this by measuring various aspects of touch and pressure. This information is sent to the machine controlling the electrodes, enabling them to fire in the correct stimulation pattern.
The acid test for a prosthesis is how well it handles delicate objects like soft fruit. For example, with tactile feedback, the men could successfully pull the stem from a cherry using their prosthesis without crushing the fruit or letting it drop. “When the tactile feedback was off you make a lot of juice,” said one volunteer.
There was an unexpected health benefit, too. Before trialling the prostheses, both men had phantom limb pain – one described it as like “a nail being driven through my thumb”, the other said it often felt like his fist was being squeezed in a vice. Over the two-year testing period, the pain diminished and eventually disappeared. In addition, although the men had to leave their sense of touch in the lab, their electrode cuffs stayed in place. Each time they went back to the lab, the reconnected cuff produced consistent sensations, even though they had led active lives doing things like chopping wood and doing DIY in between sessions (Science Translational Medicine, doi.org/v8j).
Providing a realistic sense of touch to a person who has lost a limb is a remarkable achievement. Doing the same for someone who has been paralysed by spinal cord injury is more difficult. A spinal injury prevents nerves from talking to the brain, so communications between the brain and prosthetic limbs must bypass these nerves. In 2012, a team at the University of Pittsburgh, Pennsylvania, implanted a device in the brain of a woman paralysed from the neck down, to allow her to control the world’s most sophisticated prosthetic arm, using just her mind.
The arm has 26 joints, controlled by 17 motors, and numerous sensors that can produce a high level of tactile information as the arm interacts with objects. The problem so far, says Michael McLoughlin of the Johns Hopkins University Applied Physics Laboratory, who helped develop the arm, is that only between 10 and 20 per cent of its capabilities can be exploited, because of limitations in the devices that transmit information to and from the brain. “It’s like having a smart phone and only being able to send email,” says McLoughlin. He hopes to collaborate with people like Tyler, to eventually utilise the arm’s full range of motor and sensory capabilities.
One problem in providing sensory feedback is how to read the brainwaves from one area of the brain that signal the wearer’s intention to move – in order to trigger movement in the prosthesis – and simultaneously stimulate another brain area to reproduce the sensation of touch (see diagram). That’s because electrical activity from the stimulating electrodes can interfere with the ones doing the recording.
This month, Christian Klaes at the California Institute of Technology in Pasadena and his colleagues provided a solution. By filtering out the interference from electrodes, they have developed a brain implant that could conceivably allow a paralysed person to move a robotic limb and perceive touch at the same time. Monkeys with the implant were able to control a digital version of a prosthetic arm and distinguish between two hidden objects by their textures (Journal of Neural Engineering, doi.org/v2n).
The only way to find out exactly what sensations the monkeys are getting is to see what happens in people – and Klaes aims to do this next year. He has already identified a suitable candidate and hopes to be able to do a trial in the next six months or so.
McLoughlin says that two other teams are also recruiting for candidates. “It’s a bit like using the internet in the 80s – we’re just beginning to get all this connectivity to see what will happen. When it all comes together we will achieve some tremendous advancements. It’s a very exciting time.”