25 February 2011. In Star Trek, the half-human/half-machine Borg used artificial limbs and eyes to enhance their natural abilities for purposes of galactic domination. Modern scientists, on the other hand, would like to know how to incorporate technology into the human body in order to help disabled people regain normal functions. Once pure science fiction, this idea is gradually coming closer to reality. One of the big unanswered questions is how the brain would deal with such body alterations. Intriguingly, new research suggests that not only can people accept a lifelike artificial arm as their own, but they can even accept modifications that violate the normal human body plan. In the February PLoS One, researchers led by H. Henrik Ehrsson at the Karolinska Institutet in Stockholm, Sweden, report that normal people can experience the illusion of owning a third arm without losing their feeling of ownership over their real arms. This implies that paralyzed people, for example, could incorporate an extra prosthetic arm into their body image. The feeling of owning the prosthesis might allow a person to better control such an arm, as well as having psychological benefits, the authors suggest. Nonetheless, some critical questions remain, including how well the brain could control an extra limb.
“Our study highlights that the body image is much more dynamic than was previously thought,” first author Arvid Guterstam told ARF.
In recent years, researchers have begun to crack the code for interfacing human beings and machines. Experiments conducted at Massachusetts General Hospital in Boston and at the Wadsworth Center of the New York State Department of Health in Albany, New York, have allowed people wired with electrodes to manipulate objects with their thoughts (see ARF related news story and Department of Veterans Affairs story). An even knottier problem than sending commands out from the brain to machines, however, is bringing sensory information back in. How can artificial limbs receive sensations, and be perceived by their wearer as belonging to the body?
It is not as difficult as it sounds, suggests an extensive body of research, because the wiring of the brain allows it to be tricked. In classic neurological experiments, researchers hide a volunteer’s real arm, while positioning a lifelike rubber arm in its place. When the experimenters touch the person’s hidden real hand at the same time as they touch the corresponding spot on the rubber hand, the volunteer "feels" the touch on the rubber hand. Scientists believe this illusion occurs in pre-motor areas of the brain that integrate information from different senses. In other words, because sight, touch, and proprioception (the awareness of the body’s position in space) all agree, the brain concludes that the arm it sees belongs to it (see e.g., Tsakiris et al., 2010; reviewed in Makin et al., 2008). In agreement with this, functional MRI scans reveal activity in multisensory areas of the pre-motor and posterior parietal cortex while a person experiences the rubber hand illusion (see Ehrsson et al., 2005).
Not only do people verbally report feeling ownership of the hand, brain scans corroborate it. When an experimenter appears to threaten the rubber hand by holding a knife near it, areas of the brain associated with anxiety and body awareness, i.e., the insula and anterior cingulate cortex, light up just as they do when a real hand is threatened (see Ehrsson et al., 2007). The higher the activity in multisensory areas that produce the illusion, the higher the corresponding threat response.
Ehrsson’s group was interested in testing the limits of this illusion, in particular, whether multiple limbs could be claimed. In previous work, they had found that people could simultaneously claim ownership of two rubber hands while their real hand remained hidden, suggesting the human body image is mutable (see Ehrsson, 2009). It was unclear how the presence of a real hand would affect the illusion, however. This is an important question, because many people who might benefit from an artificial limb, such as paralyzed people or people with amyotrophic lateral sclerosis, still have their real arms. To answer this question, Guterstam devised a novel experimental paradigm where volunteers placed their right arms on a table, and the researchers then placed a rubber arm next to it, but closer to the body midline. A cloth covered the volunteers from shoulder to forearm to make the visual illusion more convincing. The experimenters simultaneously stroked a small brush on the same spots of both hands for one to two minutes, while the volunteer watched. In a control condition, the experimenter stroked the two hands out of sync, which has been shown to break the illusion.
Guterstam and colleagues evaluated the success of the illusion by using questionnaires that contained statements such as, “I felt the touch of the brush on the rubber hand,” as well as statements designed to control for the person’s level of suggestibility. Participants rated how strongly they agreed with each statement. In addition, after some brushing sessions, the researchers held a knife over the rubber or real hand and measured the volunteer’s skin conductance with electrodes. Increased sweat indicated fear and provided a physiological gauge of the feeling of ownership. The results showed that participants could indeed claim ownership of two hands. Volunteers reported feeling touches on both the real and the artificial hand, feeling like the rubber hand was their own, and feeling like they had two right hands; these impressions were significantly stronger after synchronous brushing than in the control, asynchronous condition. Skin conductance responses bore this out, with more sweating when the rubber hand was threatened after a synchronous brushing session than an asynchronous one.
The authors found constraints on the illusion, as well. The phenomenon did not work if the rubber arm was rotated 180 degrees or was replaced with a rubber foot. In other words, the visual and positional information had to be correct in order for the fake arm to fool the brain. This agrees with previous studies of the classic rubber hand illusion.
Finally, the authors contrasted their paradigm with the classic illusion. They hid the volunteers’ real hands behind a partition in one set of experiments, and directly compared the results to the two-hand experiment. In the one-handed condition, the volunteers reported feeling a single touch on the rubber hand. In the two-handed setup, by contrast, people said they felt a duplicated touch on both of the hands in front of them, creating the perception of an extra limb. When participants could see both real and rubber hands, they reported a greater sense of having two right hands, which included less disowning of their real hand, as well as weaker ownership of the rubber hand, than in the one-hand setup.
In previous work, Ehrsson’s group has shown that the classic rubber hand illusion can help amputees feel like an artificial limb belongs to them (see Ehrsson et al., 2008; Rosén et al., 2009). Most amputees perceive a phantom limb, Guterstam told ARF, and a touch on a specific spot of their stump will feel, for example, like a touch on the index finger of their phantom arm. Therefore, experimenters can map out the parts of the phantom arm on the stump, and by touching the appropriate spot when a prosthetic arm is touched, elicit the illusion that the robot arm can feel. In the future, a prosthesis might be designed that could feed a touch forward to the appropriate area of the stump, Guterstam suggested. The new work suggests that the same illusion could work for people using an extra robotic arm by coordinating touches on the real and robot hands.
A significant unanswered question about extra limbs is whether the brain could learn to send motor commands to two right arms. This may be trickier than perceiving touch on a fake hand. In one recent study, for example, participants viewed a video image that made it appear that they had two left hands. When their hidden real hand and the video hands were touched in synchrony, participants felt ownership of both fake hands. When participants were asked to reach for something with their left hand, however, they made reaching errors consistent with ownership of only one of the fake hands, suggesting that only one arm at a time can be controlled by the brain (see Newport et al., 2010). Guterstam told ARF that one of their next priorities is to find out if volunteers can control false arms independently without breaking the illusion of ownership. That ability would be crucial for a useful prosthesis. Guterstam said they would also like to do brain scans of participants to see where in the brain the perception of a third arm occurs.—Madolyn Bowman Rogers.
Guterstam A, Petkova VI, Ehrsson HH. The illusion of owning a third arm. PLoS One. 2011 Feb 23;6(2). Abstract