Notice: Undefined offset: 3037 in /var/www/ on line 5961
Phantom touch - Scholarpedia

Phantom touch

From Scholarpedia
Vilayanur S. Ramachandran and David Brang (2009), Scholarpedia, 4(10):8244. doi:10.4249/scholarpedia.8244 revision #150506 [link to/cite this article]
Jump to: navigation, search
Post-publication activity

Curator: Vilayanur S. Ramachandran

Phantom Touch is continued experience of sensations and presence of a missing limb often occurring after amputation; it is frequently referred to as a phantom limb.



Figure 1: A) Normal 'homunculus' map showing physical representation of primary somatosensory cortex. B) After amputation of the hand, the 'face' area of the map invades the former 'hand' territory

About 95% of amputees experience phantoms which often emerge immediately after amputation but sometimes after weeks or months. In roughly two thirds of patients the phantom is extremely painful. Phantoms are most commonly seen after limb amputation but can occur also occur for other body parts (e.g. phantom breasts, phantom uterus, and phantom appendix). After amputation of the penis, many patients even experience a phantom penis and phantom erections (intriguingly some otherwise intact individuals also report having mainly phantom erections; S.M. Anstis, personal communication).

Although known since antiquity, phantom limbs were first described scientifically by Silas Weir-Mitchell (1872). Since then there have been hundreds of case studies reported in the medical literature (Sunderland, 1972; Riddock, 1941; Melzack, 1992) but very few systematic experiments. The current era of experimental work on human patients was inspired, in part, by animal experiments (Kaas and Florence, 1996; Jenkins, et al., 1990). The combined use of systematic psychophysics and brain imaging has allowed researchers to link neurophysiological experiments in animals with perceptual phenomenology in humans (Ramachandran, Rogers-Ramachandran, and Stewart, 1992; Ramachandran and Hirstein, 1998).

Remapping/Referral of Sensations

A few weeks after amputation of an arm, sensory stimuli applied to the ipsilateral face are experienced by the patient as arising from the missing (phantom) arm. There is often a highly specific topographically organized map of the hand on the face (Figure 2) with clearly delineated digits.

Figure 2: Topographically organized map of sensations referred from face to arm patient DS; numbers indicate digits.

This referral of sensations is possibly caused by reorganization of somatosensory maps in the brain (Figure 1). The entire right side of the body is mapped onto the postcentral gyrus of the left hemisphere, known since Penfield, and the map is systematic except for the face being directly below the hand rather than near the neck. After arm amputation the sensory input from the face, which normally projects only to the face area, now “invades” the vacated territory corresponding to the denervated hand territory. As a result, stimuli applied to the face now activate the hand region of the brain and are therefore interpreted by higher brain centers as arising from the missing phantom hand (the “remapping hypothesis”). A second map of referred sensations is often seen on the arm proximal to the amputation. This is probably caused by cross-activation of the hand area of cortex by afferents from the upper arm which normally project only to the upper arm region of cortex (Ramachandran and Hirstein, 1998; Merzenich et al., 1984). We confirmed these conjectures using MEG; a noninvasive brain imaging technique (Yang et al., 1994).

The map of referred sensations in phantom touch is modality specific; warmth on the face elicits warmth in the phantom thumb, and the same follows for cold and vibration. As touch, warmth, cold, and vibration are separately remapped in separate brain regions, the remapping must then be modality specific. While it is possible that some of this occurs in the thalamus, strokes damaging the touch fibers from the thalamic hand representation to the cortical hand map can result in sensations that are referred from face to hand, suggesting that cortical remapping is sufficient to cause the phenomenon.

Other predictions from the hypothesis were also confirmed. If the trigeminal nerve innervating the face is cut, touching the hand evokes referred sensations in the face in a topographically organized manner (Clarke et al., 1996). Amputation of a finger results in referral of sensations from adjacent fingers (Ramachandran and Hirstein, 1998) and sometimes a representation of a single finger is seen on the face (Aglioti et al., 1994). Intensity of phantom pain correlates well with the degree of remapping as explored with brain imaging, suggesting that the remapping is one of the main causes of phantom pain (Flor et al., 1995).

The remapping of hand to face possibly involves the sprouting of new axon terminals as well as the “unmasking" of pre-existing connections (Florence et al., 1998; for a review see Buonomano and Merzenich, 1998). Recent evidence for the latter hypothesis comes from the result of left ear cold caloric irrigation which is known to stimulate the vestibular cortex- adjacent to posterior insula – and probably superior parietal lobule (SPL) as well. After caloric irrigation, the patient reports the temporary reduction in the magnitude of referred sensations (Ramachandran and Azoulai, 2007), demonstrating the connections responsible for referred sensations can be inhibited or unmasked on small time scales (within a matter of hours). These findings further highlight labile nature of connections within the somatosensory cortex.

The phantom usually has a “habitual” position e.g. supinated. Remarkably a pronation of the phantom (in some subjects) also leads to a small but systematic shift in the topography of the map on the face (and proximal to the stump) along the same direction as the movement of the phantom (Ramachandran and Hirstein, 1998; Ramachandran, Brang, and McGeoch, in review), attesting to the extraordinary malleability of neural connections that determine topography in the brain. These shifts are probably caused by changes in body image in the superior temporal sulcus (STS) produced by reafference from motor commands to the phantom. It is conceivable that the signals also feedback to S1/S2 affecting topography.

In a recent study a normal volunteer placed her intact arm near the patient's phantom limb (but not overlapping the phantom). If the patient watched the volunteers hand being stroked or rubbed vigorously, he felt the sensations in his own hand (Ramachandran and Rogers-Ramachandran, 2008). This curious observation can be explained in terms of the activity of mirror neurons (Rizzolati et al., 2006). When you are touched, sensory neurons are activated in your brain’s somatosensory cortex. It has been discovered that a subset of these neurons called “mirror neurons” will fire even if you watch someone else being touched – as if the neuron was “putting you in the other person's shoes” or ‘empathizing’ with the touch delivered to the other person. But if your sensory mirror neurons fire when you watch someone else being touched why don't you literally feel her touch? This is presumably because the absence of touch signals from your skin sends a null signal that vetoes one of the outputs of the mirror neurons. If the arm were to be amputated then you would indeed quite literally feel the other person’s sensations in your phantom. This hypothesis would explain why touching another person elicits phantom sensations in the patient. Given their ability to dissolve the barrier between self and others we have dubbed these neurons “Gandhi neurons”.

In one case, suddenly “stabbing” the student volunteer's hand, the patient winced in pain and “withdrew" the phantom. This was instantly relieved by the patient simply watching the students intact hand being rubbed; an observation that might have therapeutic implications and give new meaning to the word “empathy”.

Genetic Template for Body Image

Some subjects report a phantom arm even if their arm has been missing from birth; suggesting that in spite of its extreme malleability there must also be a genetic scaffolding for body image (La Croix et al., 1992). The same might be true for transgender female to male; they often report having had a phantom penis from early childhood (Ramachandran and McGeoch, 2008). We postulate that this genetically specified scaffolding is present in the right SPL.

A “mirror image” of phantom limbs may be the curious syndrome known as apotemnophilia. In this condition a person who is otherwise completely normal experiences an intense desire to have a specific arm or leg amputated; a desire that begins in early childhood. Using MEG (magnetoencephalography) we were able to show that these patients have a body part missing from the body representation in SPL (McGeoch, Brang, and Ramachandran, in review).

Visual Feedback

Amputees usually report feeling movements in the phantom (“It's patting my brother's shoulder" etc.). When you send a command to move your arm, a copy of the command originating in the motor and pre-motor cortex is sent to the parietal cortex where you monitor these commands, combining them with visual and proprioceptive feedback to construct your body image. The monitoring of commands continues to occur even after amputation, “fooling” the brain into thinking that the phantom is moving.

In some patients the arm had been paralyzed and painful for a few months due to peripheral nerve injury. If the arm is then amputated the paralysis is “carried over” into the phantom; a phenomenon that we dubbed “learned paralysis.” We speculated that the continued absence of visual feedback signals that the motor commands are being obeyed causes the brain to learn that the arm is paralyzed and this “learned paralysis” persists in the phantom.

Figure 3: Mirror visual feedback (MVF) for phantom pain was the first demonstration that "real" somatic pain caused by cerebral changes can be modulated rapidly or eliminated by visual feedback. Such mirror feedback can also cause shrinkage of the phantom and corresponding shrinkage of pain. Similar reduction of pain has also been shown using other optical techniques and virtual reality.

The phantom is often fixed in a very painful position. You can prop up a mirror vertically (parasagittally) in front of the patient, and have him look into the mirror so that the reflection of his normal hand is superimposed optically on the felt position of the phantom ( Figure 3). This creates the illusion that the phantom arm has been resurrected and if the patient sends motor commands to make bilaterally symmetrical hand movements, the phantom appears to obey the commands. This restores the visuo-motor loop and alleviates pain (Ramachandran and Hirstein, 1998; Tsao et al., 2008) by eliminating the discrepancy that is thought to cause phantom pain (Harris, 1999). In some cases the entire phantom limb itself disappears – along with the pain (Ramachandran and Hirstein, 1998).

The efficacy of MVF for stroke rehabilitation in many patients has also now been demonstrated in a number of placebo controlled trials, following a pilot study we published in the Lancet (Altschuler, 1999; Figure 4).

Figure 4: Blind placebo controlled cross-over study demonstrating the stroking efficacy of MVF on phantom pain. The score on the visual-analogue scale ranges from 0 to 100, with higher scores indicating a greater severity of pain. Reproduced from Chan el al., 2007

These sorts of experiments suggested that there may also be a “learned paralysis" component (in addition to permanent damage to motor pathways) to the paralysis following stroke and that this could be overcome with visual feedback provided by mirrors or virtual reality (Ramachandran, 1992). There is considerable clinical evidence that this is the case (Altschuler et al., 1999; Yavuzer et al., 2008). Other neurological disorders such as complex regional pain syndrome (RSD) have also been treated successfully using the procedure (McCabe et al., 2003).

McCabe et al.’s observations are of special interest. They had patients viewing the reflection of the normal limb superimposed optically on the dystrophic limb. The mirror visual feedback procedure not only reduced pain but produced measurable changes in the temperature of the abnormal limb. This finding, that illusory visual feedback from the location of the dystrophic arm can produce instant physiological changes, cannot be “faked” and is a convincing example of genuine “mind – body” interactions.

Ordinarily our sense of having an arm arises from at least four sources 1) monitoring of feedforward or corollary discharge of commands sent from motor centers to the arm 2) Proprioceptive feedback arising from muscles and joint 3) visual feedback 4) a genetic scaffolding of ones body image. The interplay of the same sources of signals (or lack thereof) must be involved in the genesis of phantom limbs and phantom pain. In addition to pain signals arising from stump neuromas, the discrepancies between these various sources of information may itself be contributing to phantom pain and removing the discrepancies (as with the mirror) seems to relieve the pain.

Illusions of body image can also be produced viewing yourself in a “non-reversing" mirror constructed out of two mirrors with their reflecting surfaces facing each other at right angles. If you wink or move your hands randomly you can get a tingling sensation in your arms and (rarely) a momentary out-of-body experience; especially if you wear heavy makeup (or a mask) to disguise your appearance. We tested one naïve subject who noted, without prompting, that for ten minutes after she left the lab she continued to feel eerily disembodied as if she had “left her body behind” in the experimental set up. Similar experiences occur if subjects look through a half-silvered mirror to superimpose their reflection on a mask seen through the mirror. A momentary “fusion" of selves can occur (Ramachandran and Hirstein. 1998).

An observation made by Gawande (2008) is noteworthy. His patient had long suffered from a painfully “swollen” phantom. Looking at the reflection of the normal hand superposed on the phantom caused it to shrink instantly causing the pain to “shrink" as well.


During the last decade the study of phantom limbs has moved from the obscurantism of clinical phenomenology to the era of experimental science. In addition to providing insights into how the brain constructs body image, the study of phantom limbs has broader theoretical and clinical implications The idea that the adult brain consists of independent modules that are largely autonomous and fixed by genes has been replaced by the idea that the so-called modules are in a state of constant dynamic equilibrium with each other and with the sensory input. Neurological dysfunction is caused just as often by shifts in these equilibria as by permanent anatomical damage and if so pushing a reset button can restore the equilibrium; a radical concept in the rehabilitation of neurological patients. These new approaches to treatment should, however, be regarded as complimenting rather than replacing conventional therapies.


  • Aglioti, S; Smania, N; Atzei, A and Berlucchi, G (1997). Spatiotemporal properties of the pattern of evoked sensations in a left index finger amputee patient. Behavioral Neurology 111(5): 867-872.
  • Altschuler, E; Wisdom, S; Stone, L; Foster, C and Ramachandran, V S (1999). Rehabilitation of hemiparesis after stroke with a mirror. Lancet 353: 2035-2036.
  • Blakemore, S J; Bristow, D; Bird, G; Frith, C and Ward J. (2005). Somatosensory activations during the observation of touch and a case of vision-touch synaesthesia. Brain 128(7): 1571-1583.
  • Buonomano, D V and Merzenich, M M (1998). Cortical plasticity: From synapses to maps. Annual Review of Neuroscience 21(1): 149-186.
  • Chan, B L et al. (2007). Mirror therapy for phantom limb pain. New England Journal of Medicine 22: 2206–2207.
  • Clarke, S; Regali, L; Janser, R C; Assal, G and De Tribolet, N (1996). Phantom face. Neuroreport 7: 2853-2857.
  • Flor, H et al. (1995). Phantom-limb pain as a perceptual correlate of cortical reorganization following arm amputation. Nature 375: 482–484.
  • Florence, S L; Taub, H B and Kaas, J H (1998). Large-scale sprouting of cortical connections after peripheral injury in adult macaque monkeys. Science 282(5391): 1117-1121.
  • Gawande, A (2008, June 30). Annals of medicine: The itch. New Yorker 58-64.
  • Harris, A J (1999). Cortical origin of pathological pain. Lancet 354: 1464-1466.
  • Jenkins, W M; Merzenich, M; Och, M T; Allard, T and Guic-Robles, E (1990). Functional reorganization of primary somatosensory cortex in adult owl monkeys alters behaviorally controlled tactile stimulation. Journal of Neurophysiology 63: 82-104.
  • Kaas, J H and Florence, S L (1996). Brain reorganization and experience. Peabody Journal of Education 71(4): 152-167.
  • Keysers, C et al. (2004). A touching sight SII/PV activation during the observation and experience of touch. Neuron 42(2): 335-346.
  • La Croix, R; Melack, D and Mitchell, N (1992). Multiple phantom limbs in a child. Cortex 28: 503-507.
  • McCabe, C S et al. (2003). A controlled pilot study of the utility of mirror visual feedback in the treatment of complex regional pain syndrome (type 1). Rheumatology 42: 97-101.
  • Melzack, R (1992). Phantom limbs. Scientific American 266: 120-126.
  • Merzenich, M M et al. (1984). Somatosensory map changes following digit amputation in adult monkeys. Journal of Comparative Neurology 224: 591–605.
  • Ramachandran, V S; Brang, D and McGeoch, P M (in review). Dynamic reorganization of referred sensations caused by volitional movements of phantom limbs.
  • Ramachandran, V S; Brang, D and McGeoch, P M (in review). Optical shrinkage using mirror visual feedback (MVF) reduces phantom pain.
  • Ramachandran, V S and McGeoch, P (2008). Phantom penises in female to male transsexuals. Journal of Consciousness Studies 15.
  • Ramachandran, V S; Rogers-Ramachandran, D and Stewart, M (1992). Perceptual correlates of massive cortical reorganization. Science 258: 1159-1160.
  • Ramachandran, V S and Hirstein, W (1998). Perception of phantom limbs. Brain 121: 1603-1630.
  • Ramachandran, V S and Azoulai, S. Caloric stimulation modulates Phantom Limbs. Annual meeting of the Psychonomics Society.
  • Ramachandran, V S and Rogers-Ramachandran, D (2008). Sensations referred to a patient's phantom arm from another subjects intact arm: Perceptual correlates of mirror neurons. Medical Hypotheses 70: 1233-1234.
  • Ramachandran, V S; McGeoch, P M and Brang, D (2008). Society for Neuroscience abstracts.
  • Riddoch, G (1941). Phantom limbs and body shape. Brain 64: 197.
  • Rizzolatti, G; Fogassi, L and Gallese, V (2006). Mirrors in the mind. Scientific American 295(5): 54-61.
  • Sunderland, S (1972). Nerves and Nerve Injury. Edinburgh: Churchill Livingstone.
  • Yang, T T et al. (1994). Sensory maps in the human brain. Nature 368: 592-593.
  • Yavuzer, G et al. (2008). Mirror therapy improves hand function in subacute stroke: A randomized controlled trial. Archives of Physical Medicine and Rehabilitation 89: 393-398.

Internal references

Personal tools

Focal areas