Original articleBrain generators of laser-evoked potentials: from dipoles to functional significanceGénérateurs corticaux des potentiels évoqués par laser : des dipoles à la signification fonctionnelle
Introduction
Since their introduction by Carmon and his colleagues in the mid-1970s [16], [69], laser-evoked brain potentials (LEPs), and more recently laser-evoked magnetic fields (LEFs), have progressively found their place in both the clinical evaluation of patients with pain, and the investigation of cortical mechanisms sub serving the conscious, normal or abnormal pain sensation. The first decade of LEP utilisation was mainly devoted to demonstrating the selective action of laser stimuli on nociceptive skin afferents (see Plaghki et al., this volume), as well as the potential use of scalp responses in assessing damage to nociceptive pathways (see Treede et al., this volume). These aspects of LEP research were conducted using a very limited number of recording electrodes (often less than five), as they did not require a precise knowledge of the brain generators of scalp responses. The anatomical regions contributing to the responses recorded from the scalp have been progressively disclosed during the past 10 years, mainly using multi-channel recordings (20–128 electrodes or magnetic sensors), followed by source localisation methods (dipolar modelling) that use mathematical models of both the bio-electrical generators and the volume conductor where they lie. Dipolar modelling estimates the site and the orientation of intracerebral sources by comparing the scalp distribution of electromagnetic fields predicted by a given intracranial generator (modelled as a current dipole) and the actual distribution of these fields over the scalp. The solution is iteratively optimised until the predicted and the observed spatio-temporal fields are (ideally) coincident (for comprehensive and critical reviews see [43], [91], [94]). The advent of magnetoencephalographic (MEG) techniques capable to resolve overlapping sources near the scalp surface (which are often lumped in EEG recordings), has contributed precious data to the disclosure of early generators from SI and SII. On the other hand, EEG recordings have proved so far more powerful than MEG to demonstrate deep sources radial to the scalp surface, such as those arising from the anterior cingulate gyrus. One possible drawback of modelling studies is that investigators may be drawn towards trying similar solutions as previously described, thus creating a vicious circle where previous errors may have perpetuated. We try in this work to review critically this issue when relevant to the results, and to contrast as much as possible the reports on dipole modelling with those from comparable animal studies, when available, functional imaging and especially intracranial recordings in humans. In recent years, laser-evoked responses recorded with intracranial electrodes, either subdural or intracortical, have greatly contributed to refine our interpretation of modelling studies, favouring certain solutions to the detriment of others.
In this paper, we review and analyse critically previous EEG and MEG results on the anatomical sources of LEPs, including recent data from our laboratories, and provide some tentative hypotheses on the functional significance of generators so far disclosed.
Section snippets
Pioneering studies: suprasylvian opercular sources, including second somatic area (SII)
The three seminal modelling studies on LEP generators [9], [96], [102] were obtained using the same modelling algorithm based on spatio-temporal modelling with fixed (i.e. non-moving) dipoles (BESA; [94]). In spite of different stimulation sites (hand or temple), variable number of electrodes in each case (15 vs. 20 vs. 31), and different laser types (CO2 vs. thulium-YAG 1
Primary sensory cortex (SI)
In their pioneer dipole EEG study, Tarkka and Treede [96] suggested that a SI source was active simultaneously to, or even later than, suprasylvian areas. However, while the involvement of opercular areas proved to be a most consistent finding in subsequent work (see above), the presumed activation of SI was not reproduced in other modelling studies, either using EEG with up to 64 recording electrodes [4], [9], [19], [102], [104], or MEG with up to 74 channels [52], [72], [111], [114].
Anterior cingulate cortex
In EEG modelling studies, sources in the anterior cingulate cortex (ACC) have been almost as frequently described as those in suprasylvian regions, and this either after conventional A-delta stimuli [9], [88], [96], [102], [103] or following specific C-fibre excitation [22], [47], [73]. The latency times of ACC activation, although slightly variable across studies, lie within the window of the main biphasic vertex LEP (N2-P2) component. Although the anterior–posterior axis of the ACC is very
Other sources (insular, mesio-temporal and frontal lobe generators)
Some other sources have been occasionally described in EEG modelling and MEG studies but with much less consistency than the above-mentioned areas. In the study of Bentley et al. [4], the N1-N2 LEP responses (150–250 ms) to A-delta stimuli were not explained by opercular sources, but rather by a posterior parietal generator, followed by cingulate (400 ms) and anterior insular (800 ms) responses, without opercular contribution. The presence of insular-generated responses to A-delta inputs had
Conclusion
The cortical regions most consistently tagged as sources of scalp LERs are the suprasylvian region (parietal operculum, SII) and the anterior cingulate cortex (ACC); their activity in response to laser has been supported by intracranial recordings. Primary somatosensory sources are also probably active in response to laser, their response being largely parallel to that of opercular sources. Insular responses are also supported by intracranial recordings but have been described less consistently
Acknowledgements
This work was supported in part by a Grant of the CNP Foundation against pain (Fondation CNP, “Etude électrophysiologique et clinimétrique des facteurs de survenue de la douleur neuropathique”) to Luis Garcia-Larrea, and by a Grant of the UPSA Institute for pain to Maud Frot (“Enregistrements intracorticaux des réponses aux stimulations nociceptives”).
References (116)
- et al.
Magnetoencephalographic responses to painful impact stimulation
Brain Res
(1999) - et al.
Source localisation of 62-electrode human laser pain evoked potential data using a realistic head model
Int J Psychophysiol
(2001) - et al.
Anatomical localization and intra-subject reproducibility of laser evoked potential source in cingulate cortex, using a realistic head model
Clin Neurophysiol
(2002) - et al.
Subcortical structures involved in pain processing: evidence from single-trial fMRI
Pain
(2002) - et al.
Brain electrical source analysis of laser evoked potentials in response to painful trigeminal nerve stimulation
Electroencephalogr Clin Neurophysiol
(1995) - et al.
Cortex areas involved in the processing of normal and altered pain
Prog Brain Res
(2000) - et al.
Brain systems mediating aversive conditioning: an event-related fMRI study
Neuron
(1998) - et al.
Somatotopy of human hand somato-sensory cortex revealed by dipole source analysis of early somatosensory evoked potentials and 3D-NMR tomography
Electroenceph Clin Neurophysiol
(1995) - et al.
Laser-evoked potentials in human pain I. Use and possible misuse
Pain Forum
(1998) - et al.
The assessment of pain by cerebral evoked potentials
Pain
(1983)