[Limitations of and new options with neurophysiological methods in migraine research and diagnosis : Electroencephalographicmethods.].

In parallel with the development of the various electrophysiological methods now available, the clinical syndrome of migraine has been repeatedly analyzed with the aid of these tools, in attempts to track its course with reference to electrophysiological parameters. A survey of the international literature reveals the following EEG findings. (1) In patients with different forms of headache without structural lesions the distribution of normal and abnormal EEG recordings during headachefree intervals resembles that seen among healthy subjects. (2) In common migraine, nonfocal and nonspecific abnormalities are reflected in up to 65% of all EEG recordings during the migraine-free interval. (3) Focal EEG abnormalities during a migraine attack are relatively rare, generally occurring only during migraines with transient focal neurological disturbances. (4) It is very rare for there to be permanent focal changes that are apparent during both pain-free intervals and migraine attacks. In the presence of such changes very careful diagnosis is necessary to exclude symptomatic lesions. (5) In up to 53% of migraineurs a few minutes' hyperventilation leads to generalized slow, wide-amplitude, sometimes episodic waves in the EEG. Similar changes also occur in healthy subjects, but are less common, less intense and of shorter duration. (6) There is a statistically significant correlation between the occurrence of a migraine and EEG activation by photic stimulation at frequencies over 20 Hz. (7) No unequivocal results obtained in large series of patients with train mapping are yet available.