Neuromagnetic source localization of auditory evoked fields and intracerebral evoked potentials: a comparison of data in the same patients
Introduction
Different components of an auditory evoked potential (AEP) can be isolated and described as a function of latency, scalp topography, and the perceptual/cognitive process to which they are associated. Each component underlies a different aspect of auditory processing, beginning in the cochlea and terminating in the auditory cortex. The earliest components (<20 ms), corresponding to activity in the thalamo-cortical volley, are not reliably identifiable from the scalp. Several middle-latency potentials, however, can be identified from the scalp: Na (or Nam in the MEG literature) at 19 ms, Pa (or Pam) at 30 ms, Nb (or Nbm) at 40 ms, and Pb or P1 (Pbm or P1m) at 50 ms. Neuromagnetic data show that, while the Pam component is consistently observed in normal subjects, the other 3 are less reliable, occurring more or less frequently according to the study (Pelizzone et al., 1987, Reite et al., 1988, Scherg et al., 1989, Pantev et al., 1990, Pantev et al., 1993, Mäkelä et al., 1994). Pam's source has been localized in a region more anterior and more medial to that generating N1m (a large, robust late auditory evoked response peaking at 100 ms), suggesting that these components are generated in different regions of the auditory cortex. While P1, however, is thought to be generated in the primary auditory cortex (Huotilainen et al., 1998, Reite et al., 1988), neuromagnetic techniques fail to provide much data on the underlying neural sources of this and other middle-latency components because of the small amplitude that these components often have. Their sources may be localized more reliably from intracerebral recordings using depth electrodes implanted in the auditory cortex for the pre-surgical evaluation of medically intractable epilepsy.
In order to accurately localize and delineate the epileptic zone in the epileptic subject, pre-surgical evaluation (stereo-electro-encephalography (SEEG)) requires the stereotaxic implantation of multiple depth electrodes (Bancaud et al., 1965). Because the size and location of each cortectomy is different for each patient, it is important to have high spatio-temporal resolution of ictal and inter-ictal electrical activity. The superior temporal gyrus (STG) is a key structure in lateral and mesial–lateral temporal lobe epilepsy and is frequently evaluated intracerebrally (Bancaud and Talairach, 1992). Situated in the STG, the auditory cortex consists of several different anatomical structures. It includes the transverse gyri of Hschl (the dorso-postero-medial part of which corresponds to the primary auditory cortex; the lateral part of which corresponds to the secondary areas (Liégeois-Chauvel et al., 1991)). The secondary areas continue caudally onto the planum temporale (PT) and extend ventrally above and slightly below the superior temporal sulcus.
Previous data suggest the existence of at least 7 middle-latency generators underlying their respective components at 13, 16, 30, 50, 60, 75, and 100 ms. Their sources are localized in different regions along the Heschl's gyrus (HG). N/P 16 and 30 originate from the most medial part of the primary cortex (i.e. the dorso-postero-medial part of the HG); N/P 50 (P1) is generated in the lateral part of the primary cortex; N/P 60 and 75 originate from the intermediate or lateral part of the HG (secondary cortex); and the N100 has at least two generators – one in the lateral part of HG and one in the PT (Liégeois-Chauvel et al., 1991, Liégeois-Chauvel et al., 1994). With respect to the generator of this latter component, much of the previous data has been discordant with some authors suggesting its localization in the primary auditory cortex (Liégeois-Chauvel et al., 1994, Huotilainen et al., 1998).
Intracerebral recordings provide direct and accurate information on the localization of a component's generator. Scalp recordings, on the other hand, provide at best an estimate of a generator's localization, which must be calculated and which cannot be measured directly. Localizing these sources accurately requires the modeling of intracerebral sources and of the head, and the choice of the model affects spatial and temporal accuracy. While this presents a challenge for electroencephalography (EEG), magneto-encephalography (MEG) is a non-invasive technique, which offers better spatial resolution owing to highly sensitive captors and to the fact that it is less subject to interference from the skull than is EEG (Wikswo et al., 1993, Gevins, 1996). Recently, we had the opportunity to record and study neuromagnetic scalp and intracerebral AEPs in the same patients, making a comparison between magnetic source localization and source localization using depth electrodes possible.
The aim of the present study was to compare intracerebral and scalp auditory evoked responses and their respective source localizations. Such an endeavor would help us better appreciate the resolving power of the spatio-temporal algorithm currently used for MEG in our laboratory and could ultimately help assess the localization capacity and utility of non-invasive techniques in clinical settings.
Section snippets
Subjects
Recordings were carried out in 4 right-handed adult patients (3 males; one female), aged 21–43 years, with medically intractable partial seizures. In no patient was the epileptogenic foyer localized in the STG. During each recording session, subjects were laid comfortably in a bed and were instructed to remain attentive. In 3 patients, the right hemisphere was explored; in one patient, the left hemisphere was explored.
All subjects gave informed consent for all of the examinations and
Results
Fig. 2 (top) shows AEPs recorded in Case 1 from 9 consecutive leads along electrode H′ in the medial part of left HG (3 leads (H′1–3)) and PT (5 leads (H′4–8)) in response to the click stimuli. Lead H′1 recorded short-latency evoked responses (N15/P20/N30) characteristic of responses from the primary auditory cortex, followed by middle-latency components (P45/N60) and a slow-wave peaking around 90 ms. The amplitudes of the early components were smaller on leads H′2 and H′3 and disappeared
Discussion
The localization of generators from the SEEG data yielded similar results to those cited previously (Liégeois-Chauvel et al., 1994). In Case 1 (using click stimuli), most of the middle-latency potentials (13–50 ms post-stimulus onset) previously recorded in the primary auditory cortex (Liégeois-Chauvel et al., 1991, Liégeois-Chauvel et al., 1994) were observed. These responses show a polarity inversion compared to EEG surface responses (Streletz et al., 1977, Özdamar and Kraus, 1983, Cacace et
Acknowledgements
We would like to thank J.M. Scarabin who was responsible for the implantation of electrodes and surgical intervention, J.M. Badier whose critical comments on a previous version of this paper were of great aid, P. Marquis for his technical assistance, and K. Giraud for revising the English version.
References (34)
- et al.
Human middle latency auditory evoked potentials: vertex and temporal components
Electroenceph clin Neurophysiol
(1990) - et al.
Sequential mapping favours the hypothesis of distinct generators for Na and Pa middle-latency auditory evoked potentials
Electroenceph clin Neurophysiol
(1988) - et al.
A study of the auditory evoked magnetic fields of the human brain
Electroenceph clin Neurophysiol
(1980) - et al.
Combined mapping of human auditory EEG and MEG responses
Electroenceph clin Neurophysiol
(1998) - et al.
Evoked potential recorded from the auditory cortex in man: evaluation and topography of the middle latency components
Electroenceph clin Neurophysiol
(1994) - et al.
Bi-hemispheric study of single trial MEG signals of the human auditory cortex
Electroenceph clin Neurophysiol
(1998) - et al.
Whole-head mapping of middle latency auditory evoked magnetic fields
Electroenceph clin Neurophysiol
(1994) - et al.
Identification of sources of brain neuronal activity with high spatiotemporal resolution through combination of neuromagnetic source localization and magnetic resonance imaging
Electroenceph clin Neurophysiol
(1990) - et al.
Relationship of transient and steady-state auditory evoked fields
Electroenceph clin Neurophysiol
(1993) - et al.
Specific tonotopic organizations of different areas of the human auditory cortex revealed by simultaneous magnetic and electric recordings
Electroenceph clin Neurophysiol
(1995)