Phantoms in the brain

BMJ 1999; 319 doi: https://doi.org/10.1136/bmj.319.7210.587 (Published 04 September 1999) Cite this as: BMJ 1999;319:587
  1. Peter W Halligan, MRC senior research fellow,
  2. Adam Zeman, consultant neurologist,
  3. Abi Berger, science editor
  1. Department of Experimental Psychology, University of Oxford, Oxford OX1 3UD
  2. Department of Clinical Neurosciences, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU
  3. BMJ

    Question the assumption that the adult brain is “hard wired”

    After amputation many people experience vivid sensations of the body part they have lost.1 These “sensory ghosts” can arise within hours of the loss of the limb and are often painful. Such phenomena have been a mysterious part of medical lore for over a century, but recent research suggests that phantoms can teach us substantial lessons about the organisation and plasticity of the brain.

    Fig 1.
    Fig 1.

    How amputees perceive their phantoms

    (Credit: ALEXA WRIGHT)

    Embedded Image

    We stand to learn most from phantoms if we attend closely to patients' subjective reports. One innovative study, for example, has made use of digital photography to depict how amputees perceive their phantoms (fig 1) (A Wright et al, Wellcome Trust Sci Art Project, 1997). By remaining true to patients' own experiences, the researchers found it possible to document several neglected features. Patients reported a wide range of phantom sensations and many described striking changes in the phantom over time. Similar phantom sensations have been reported to occur after mastectomy2 and after stroke.3

    Similar work by Aglioti with women undergoing mastectomy found that 25% of patients experienced a phantom breast when the pinna region of the ear lobe (on the same side as themastectomy) was stimulated.2 This surprising finding suggests that these two regions might share neighbouring neural representations on thesensory homunculus in the brain.

    Several recent brain imaging experiments also lend support to the functional remappinghypothesis. Studies of arm amputation using magnetoencephalography—a technique which allows high fidelity recordings of the electrical activity of small regions of the cortex—have shown that brain areas which ordinarily represent the hand were activated when either the lower face or upper arm was touched.6 7 Functional imaging techniques, such as positron emission tomography, provide another means of establishing which brain areas are activated by specific stimuli and tasks. One recent study demonstrated that stimulation of the trunk with touch reliably activated the representation of the hand on the sensory cortex, showing that sensory input to the trunk had gained access to the “homuncular hand.”8

    Fig 2.
    Fig 2.

    The areas depicted on the right side of the face in this patient elicited the precisely localised referred sensations in the phantom hand, shown with supine position

    Given the relatively short delay between amputation and phantom experience—usually less than 24 hours—this “invasion” of neuronal connections is unlikely to result from the sprouting of new connections. A more plausible explanation is that existing dormant synapses between neighbouring cortical areas are unmasked after large scale sensory loss. However, both processes could be at work, and indeed this remapping may involve neuronal changes at both cortical and thalamic levels.9

    All of these findings question the widely held assumption within neuroscience that the primary sensory areas of the adult brain are hard wired There is now growing evidence that complex brain adaptation can occur after stroke, and functional remapping may explain the striking degree of recovery seen in some patients. It has also been shown that extensive training can produce physical changes in specific parts of the brain: Braille readers and trained musicians show highly developed brain areas responsible for touch and musical appreciation.10

    Understanding how the brain reorganises itself may eventually facilitate the treatment of patients with intractable phantom pain. Ramachandran, for example, has proposed a novel technique for the relief of the cramp-like pain caused by a clenched phantom fist. By observing the reflection of their free moving normal hand in a mirror, so that it appears in the position of the phantom, some patients report an immediate and compelling illusion whereby they are able to generate voluntary movements in the phantom hand. In some cases, the sensation of unclenching the phantom hand relieves the pain.5 The efficacy of this method will require a prospective controlled trial, but it offers the prospect of a real clinical benefit from an exciting insight into how the brain organises itself.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    View Abstract

    Sign in

    Log in through your institution