Examining the Uses of Virtual Reality in Prosthetic Rehabilitation

By Phil Stevens, MEd, CPO, FAAOP

Prosthetics patients must adapt to a variety of new realities. They place their body weight on artificial legs and reach out with robotic hands that respond to a series of contrived operational signals from their body. Their prostheses, in turn, support their body weight and lift objects around them. Is there a place, then, for virtual reality as prosthetics patients adjust to their missing limbs and adapt to external prostheses? This article focuses on the evolution of virtual-reality applications in the treatment of prosthetics patients.

Mirror, Mirror...in a Box?

Vilayanur Ramachandran, MD, PhD, director of the Center for Brain and Cognition and professor with the Psychology Department and the Neurosciences Program at the University of California, San Diego (UCSD), is credited with creating one of the earliest virtual-reality environments for the treatment of patients with amputations. In the mid 1990s, he reported on a cohort of upperlimb amputee subjects who completed various exercises using what he described as his "virtual reality box"-a box with no top into which he placed a mirror so that it divided the box vertically into two halves. He cut out two adjacent holes on one side of the box so that each hole opened into one of the two halves. Study subjects were then asked to insert their sound-side hand into the half with the reflective surface and the amputated limb into the other half, and then lean into a position so that the reflection of the sound-side hand was visible-as if it were their missing limb.

In Ramachandran's pioneering work, patients were able to send simultaneous motor commands to both limbs whereupon the phantom hand, as a reflection of the sound limb, was not only visible to the patient, but appeared to move in accordance with the patient's motor signals.

Mirror box.

Vilayanur Ramachandran's mirror box.

The value of this virtual recreation of the amputated hand is the discovery that patients have the ability to overcome the "learned paralysis" of the phantom hand. While most upperlimb amputees experience the sensation of a phantom hand following their amputations, many patients are unable to control the "movement" of the phantom limb. For some, this leads to the frustration of a "paralyzed" phantom limb. Some people experience the paralyzed limb as being in very painful postures, such as a tightly clenched fist in which the fingernails dig aggressively into the palm. Because they are unable to control the phantom limb, patients are unable to alter this position and relieve the associated pain.

Ramachandran's original case study includes descriptions of several upper-limb amputee patients who experienced these painful, frozen phantom hands in undesired postures. When their eyes were shut and there was no "virtual" phantom hand, the phantom sensations were those of a hand "frozen in a cement block," regardless of whether or not they sent motor commands to both limbs and moved their sound-side limb. When the patients opened their eyes and were able to see their virtual hand respond to motor signals, several patients described sensations of movement within the phantom hand, allowing them to unclench it and relieve the associated pain.1

Further Reflections

Mirror therapy to treat phantom pain

Navy Cmdr. Jack Tsao, MD, associate professor of neurology at the Uniformed Services University of the Health Sciences, Bethesda, Maryland, encourages Army Sgt. Nicholas Paupore, an outpatient at Walter Reed National Military Medical Center, Washington DC, to try mirror therapy to treat phantom pain in his amputated right leg. Tsao has conducted clinical trials in mirror therapy and said he hopes to advance the study to bring relief to amputees from Iraq and Afghanistan. Photograph by Donna Miles, courtesy of the U.S. Department of Defense.

Given the promise of these early observations in upper-limb amputees, it is somewhat surprising to note that it would be more than ten years before the efficacy of mirror-box therapy in reducing phantom pain would be evaluated in randomized, controlled trials and that these would be performed in the lowerlimb amputee population. The conflicting observations of the following two trials may well have been a product of their differing methodologies.

In the first trial, patients were assigned to a single treatment session in one of two conditions. The experimental group used a mirror box, similar to Ramachandran's, and completed a single treatment session that included a clearly identified series of movements with both the sound side and phantom limb while visualizing the virtual limb in the form of the reflected image. A control group performed the same series of movements, but with the mirror obscured so that they could not see their virtual limb. Only a small minority of test subjects reported phantomlimb pain prior to the treatment session, and of these, about half of the subjects in each of the two cohorts reported complete pain relief.2

In the second controlled trial, published the same year in the New England Journal of Medicine, lower-limb amputee patients with phantom-limb pain completed four weeks of daily treatment in one of three groups: (1) a mirror therapy group; (2) a covered mirror therapy group; and (3) a mental imagery of movement group. The subjects in each group used the same assigned techniques for 15 minutes a day during the four-week treatment protocol. At the conclusion of the first round of treatment, all six of the subjects assigned to the mirror therapy group reported decreased phantom-limb pain. By contrast, only one of the six subjects in the covered mirror therapy group and two of the six subjects in the mental imagery group reported a decrease in pain. Interestingly, three of the six patients in the covered mirror group and the remaining four subjects in the mental imagery group reported that their pain was worse at the conclusion of treatment.3

The authors of this second trial went on to report that nine of the 12 subjects not originally assigned to the mirror treatment group subsequently began a four-week trial of mirror therapy, at the conclusion of which all but one reported a decrease in phantom-limb pain.3 Taken together, these trials would appear to suggest that repeated exposure to the virtual images of the mirror box may be necessary to improve the likelihood of reducing phantom-limb pain symptoms.

A third, retrospective review of the effects of mirror-box therapy and its potential contraindications also bears mentioning. In this study, a rehabilitation team describes its experience integrating 30-minutes per day of mirror-box therapy for three weeks as part of its inpatient rehabilitation process following lowerlimb amputation.4 Of the 33 subjects who were retrospectively considered, only four subjects completed the full three-week mirror-box therapy treatment. Ten subjects withdrew immediately after the first session, and ten more subjects withdrew after two days of treatment. The nine remaining subjects withdrew after three to five days of treatment. The study subjects said that they withdrew from the mirror-box treatment because it made them feel dizzy, irritated, and uneasy. Importantly, there was no mention of phantom-limb pain in any of these cases, so there does not appear to have been any perceived purpose in the treatment. Thus, the routine application of mirror-box therapy for a new amputee, especially a lower-limb amputee for whom phantom-limb pain is both less common and less severe, may not be in the best interest of the patient.

Refining the False Realities

Mirror box.

A diagrammatic explanation of the mirror box. The patient places the sound-side limb into one side of the box (in this case the right hand) and the amputated limb into the other side. Due to the mirror, the patient sees a reflection of his sound-side hand where his or her missing limb would be (indicated in lower contrast). The patient thus receives artificial visual feedback that the "resurrected" limb is now moving when he or she moves the sound-side hand. Image created by Edward M. Hubbard for Wikipedia (August 10, 2006; updated July 28, 2007).

For all of its promise, there are inherent limitations to the virtual reality created within the mirror box. It presupposes that the phantom limb is located at a length comparable to the sound-side limb. Because it does not address the "telescoping" that is often experienced with phantom sensations (i.e., the perception that the phantom forearm is shortened, bringing the phantom hand closer to the elbow), it requires patients to assume and maintain a fairly fixed position and ignore the visible movements of their intact limbs. As computer-generated virtual realities have become more convincing, researchers have been able to apply a more immersive virtual reality to the treatment of phantom-limb pain.

In one such investigation, a pair of upper-limb amputees were fitted with a head-mounted virtual-reality display that showed the virtual surroundings as though from the user's normal field of vision.5 The subjects wore data gloves on their sound-side limbs with additional sensors attached to their wrists and elbow joints. The subjects then performed defined tasks, and the movements from the sound-side limb were transposed into the virtual reality created by the head-mounted display. As with Ramachandran's observations, the contrived reality felt very real to the patients and lead to some degree of relief from phantom-limb pain.

In a related attempt, researchers sought to eliminate the need for bilateral, symmetrical movements of the sound and phantom limbs by having the residual limbs themselves generate the movements that would, in turn, produce the virtual images.6 In their protocols, both upper- and lower-limb amputees with histories of phantom-limb pain were fitted with electromagnetic sensors on their residual limbs. These sensors relayed movement events that were transposed into one of two virtualreality environments designed for use on home computers (arm amputees reached for and picked up a virtual apple from a virtual table, while leg amputees stomped on the virtual pedal of a virtual bass drum).

Following a calibration and training procedure, the majority of the upper-limb amputees reported a sense of "agency" (that the virtual limb was, in fact, their limb that they were controlling on the screen) and a concomitant reduction in phantom-limb pain. The two upper-limb subjects who did not experience this sense of agency had both had nerve root avulsions five and 18 years prior to their amputations; therefore, their phantom limbs were paralyzed. It is noteworthy that these were the only subjects who did not report a reduction in phantom-limb pain.

Lower-limb amputees reported less severe pain than their upper-limb peers, characterized more by painful spasms compared to the continuous pain of the upper-limb patients. The majority of lower-limb subjects also experienced what the authors described as a "sense of immersion" into the virtual environment, which coincided with an associated reduction in pain. Interestingly, a third cohort of lower-limb subjects with no reports of active phantom-limb pain was also evaluated. None of these subjects found purpose in the virtual procedure and consequently, failed to immerse themselves into the contrived reality. This further reinforces the futility of mirror-boxtype therapies in the absence of phantom-limb pain.

Transitioning from Pain to Function

Weight-shifting video games Weight-shifting video games

Weight-shifting video games, such as the Nintendo Wii Fit, offer opportunities for balance-training therapies. Photographs by Lauren Muir, courtesy of the Holland Bloorview Kids Rehabilitation Hospital.

Several more recent studies have reported on attempts to use virtual-reality environments to improve functional performance. A recent case study describes the use of a very sophisticated virtual reality as a method of gait training for a study subject some two years after his transfemoral amputation.7 The subject was positioned on a treadmill in an elaborate domed structure. Within the dome, eight projectors were used to simulate a straight walk through a forested environment. In addition, a 24-camera motion-capture system was used to track the subject's full-body kinematics and create a real-time, full-body virtual representation of him walking in the virtual environment. A treating physical therapist was positioned behind the subject to analyze the subject's kinematics in real time and provide verbal cues to assist the subject in correcting his deviations by using the visual feedback presented from the virtual environment. Following 12 sessions, the subject demonstrated improved hip, pelvic, and trunk mechanics with reduced oxygen-consumption values.7

Similar principles were used in a much more practical approach in a recent case series in which pediatric subjects with unilateral transfemoral and unilateral Van Ness-level amputations were evaluated before and after a prescribed training regimen using a Wii gaming device.8 Citing precedence of improved motivation and compliance with such approaches, the researchers asked the subjects to play two Wii Fit games per day for 20 minutes each, four days a week for four weeks. Both games required the subjects to stand on the game system's balance platform and control their center of mass. The games included Table Tilt, in which players shift their weight to tilt a virtual table and move virtual marbles into a hole, and Tightrope Walk, in which coronal weight shifts are used to guide a virtual character across a suspended rope.

As anticipated, compliance with the regimen was quite high. The subjects with the Van Ness-level amputations presented with center of mass displacement values comparable to normal controls, and as a result experienced little improvements post intervention. By contrast, the patients with transfemoral amputations originally presented with substantial balance deficits and improved significantly following the prescribed virtual gaming. For example, the sway area experienced by these subjects in quiet standing decreased by an average of 41 percent immediately following the four-week training session. The study subjects' improved scores on the Community Balance and Mobility Scale before and after the training procedures provide further evidence of improved balance.

Reprogramming Motor Control

One of the challenges associated with upper-limb prosthetic care is for patients to learn to associate new body motions and signals with movement of their prostheses. Ottobock's MyoBoy® software has long provided a form of virtual reality as the muscle contractions of the learning patient are transformed into the movements of a virtual hand on a computer screen or the vertical position of virtual cars as the patient navigates through a series of gates.

As prosthetic arms have increased in complexity, however, more elaborate virtual-reality training techniques have been developed. The U.S. Department of Veterans Affairs (VA) has developed and reported on such attempts in the article, "Study to Optimize the DEKA Arm."9

The DEKA Arm can provide up to ten types of powered movement using elaborate control strategies that may include various foot controls, air bladders, switches, and myoelectric controls. Inherently, many of these control motions are unrelated to the ultimate prosthetic motions they generate. One example cited by the authors is the use of pronation and supination of the foot to control pronation and supination of the prosthetic wrist. Consequently, mastery of the DEKA Arm requires a great deal of motor re-training.

Researchers are using virtual reality to facilitate this motor re-training prior to the fitting of the prosthesis and to provide a staged introduction of control motions. Before subjects wear the DEKA Arm, they are fitted with the controls and exposed to a screen-based virtual reality where they see an avatar of themselves and learn to generate the movements that will control the motions of the elbow, forearm, wrist, and hand. Once fitted with the arm, patients continue to train in the virtual environment with the arm deactivated so that they can refine the new motor pathways that will convert their movements into the desired prosthetic motion.

Researchers from the U.S. Department of Defense provide an in-depth account of the experiences of a single case study learning to control the complex arm during 3.5 hours of virtualreality- based training.9 The training occurred over eight sessions. During the first seven sessions, the subject wore only his socket and its controls. During the final session, the subject wore the deactivated DEKA Arm. After these sessions, the subject proceeded with training with the actual arm in a real environment.


Following their amputations, the realities of our patients change substantially. It is therefore unsurprising to learn of the value of virtual-reality approaches in helping them to adapt to the absence of their missing limbs and improve their control of their new prosthetic devices.

Phil Stevens, MEd, CPO, FAAOP, is in clinical practice with Hanger Clinic, Salt Lake City, Utah. He can be reached at


  1. Ramachandran,V. S., and D. Rogers-Ramachandran. 1996. Synaesthesia in phantom limbs induced with mirrors. Proceedings of the Royal Society B: Biological Sciences 263:377-86.
  2. Brodie, E. E., A. Whyte, and C. A. Niven. 2007. Analgesia through the looking-glass? A randomized controlled trial investigating the effect of viewing a "virtual" limb upon phantom limb pain, sensation and movement. European Journal of Pain 11:428-36.
  3. Chan, B. L., R. Witt, A. P. Charrow, et al. 2007. Mirror therapy for phantom limb pain. The New England Journal of Medicine 357:2206-7.
  4. Casale, R., C. Damiani, and V. Rasati. 2009. Mirror therapy in the rehabilitation of lower-limb amputation: Are there any contraindications? American Journal of Physical Medicine & Rehabilitation 88:837-42.
  5. Murray, C. D, S. Pettifier, T. Howard, et al. 2007. The treatment of phantom limb pain using immersive virtual reality: Three case studies. Disability and Rehabilitation 29(18):1465-69.
  6. Cole, J., S. Crowle, G. Austwick, and D. H. Slater. 2009. Exploratory findings with virtual reality for phantom limb pain; From stump motion to agency and analgesia. Disability and Rehabilitation 31(10):846-54.
  7. Darter, B. J., and J. M. Wilken. 2011. Gait training with virtual reality-based real-time feedback: Improving gait performance following transfemoral amputation. Physical Therapy 91(9):1385-94.
  8. Andrysek, J., S. Klejman, B. Steinnagel, et al. 2012. Preliminary evaluation of a commercially available videogame system as an adjunct therapeutic intervention for improving balance among children and adolescents with lower limb amputations. Archives of Physical Medicine and Rehabilitation 93:358-66.
  9. Resnik, L., K. Etter, S. L. Klinger, and C. Kambe. 2011. Using virtual reality environment to facilitate training with advanced upper-limb prosthesis. Journal of Rehabilitation Research & Development 48(6):707-18.