Images Of Health

Can you tell your clunis from your cubitus? A benchmark for functional imaging

BMJ 2004; 329 doi: (Published 16 December 2004) Cite this as: BMJ 2004;329:1492
  1. Alison E Fisher, contract research fellow1,
  2. Gareth R Barnes, tenured research fellow (barnesgr{at},
  3. Arjan Hillebrand, tenured research fellow1,
  4. Caroline Burrow, contract research fellow1,
  5. Paul L Furlong, lecturer in psychology1,
  6. Ian E Holliday, lecturer in psychology1
  1. 1 Neurosciences Research Institute, Aston University, Birmingham B4 7ET
  1. Correspondence to: G R Barnes

    Advances in functional brain imaging have allowed the development of new investigative techniques with clinical application—ranging from presurgical mapping of eloquent cortex to identifying cortical regions involved in religious experiences. Similarly a variety of methods are available to referring physicians, ranging from metabolic measures such as functional magnetic resonance imaging and positron emission tomography to measurements based on electrical activity such as electroencephalography and magnetoencephalography. However, there are no universal benchmarks by which to judge between these methods. In this study we attempt to develop a standard for functional localisation, based on the known functional organisation of somatosensory cortex.1

    Studies have shown spatially distinct sites of brain activity in response to stimulation of various body parts.2 Generally these studies have focused on areas with large cortical representations, such as the index finger and face.3 4 We tested the limits of magnetoencephalography source localisation by stimulation of body parts, namely the clunis and the cubitus, that map to proximal and relatively poorly represented regions of somatosensory cortex.

    Participants, methods, and results

    Three participants (two men, one woman, aged 25-35 years) lay comfortably in a whole head Omega magnetoencephalograph (CTF Systems). We attached single disposable electrodes to each participant's right clunis (upper portion) and across the right cubitus. Electrical stimulation was delivered at twice the sensory threshold for each site. An experimenter sat beside each participant within the magnetically shielded room and operated an electrical switch to alternate the stimulation site when prompted by a visual cue. Fifty stimuli were delivered to each site in a boxcar design (5 seconds on, 5 seconds off). We collected magnetoencephalographic data using a 625 Hz sampling rate and averaged them for each stimulation site. We then co-registered the data with each participant's anatomical magnetic resonance image, using a surface matching approach. A single equivalent current dipole was fitted to the first evoked response peak. For each reconstructed location of brain activity, we used a Monte Carlo simulation to generate 95% confidence ellipsoids.

    What is already known on this topic

    The ability to differentiate between one's clunis and one's cubitus is a nationally accepted minimum level of professional and social performance

    What this study adds

    The same metric could be used to judge between brain imaging techniques

    The figure shows identified sites of cortical activity, and 95% confidence ellipsoids, corresponding to stimulation of the clunis and cubitus superimposed on a representative magnetic resonance image. The two cortical sites are clearly distinct, with no overlap of the 95% confidence volumes. Furthermore, the data are in good agreement with Penfield's neurosurgically established homunculus.1

    Embedded Image

    Magnetoencephalography identified sites of brain activity in the left hemisphere, including 95% confidence ellipsoids, for stimulation of right cubitus (yellow) and clunis (purple). Note the close correspondence with Penfield's homunculus


    We found that magnetoencephalography can successfully differentiate your clunis from your cubitus, despite the small cortical representation and close proximity of these areas, and despite informal behavioural observations which suggested that one of the participants (IEH) was incapable of such a distinction himself. Further work might involve the investigation of gender differences. To conclude, perhaps the most accessible and objective heuristic for the evaluation of any functional imaging technique comes in the form of a popular English idiom.5


    • We thank the reviewers, who would clearly make model subjects for this study, for their insightful comments.

    • Contributors All authors contributed equally to this work. GRB is guarantor for the study.

    • Funding The magnetoencephalograph used was jointly funded by HEFCE and the Wellcome Trust. AEF is funded by the Lord Dowding Fund for Humane Research.

    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    View Abstract