Interictal indices of temporal seizure origin.

Two studies assessed the value of temporal lobe interictal electroencephalographic (EEG) spikes and delta in indicating side of temporal epileptogenesis. The first study determined laterality of spikes/delta in awake recordings of 56 patients whose seizures all began unilaterally as proven by (1) EEG-recorded seizures and (2) > 90% improvement after lobectomy. Spikes of 52 (93%) and delta of 46 (82%) patients predominated or appeared exclusively ipsilateral to seizure origin. Neither predominated contralaterally in any patient. The second study investigated laterality of temporal seizures in a separate group of 156 patients with various side vs side spike or delta ratios on 1 to > or = 4 awake recordings. Ninety-nine of 104 patients (95%) with temporal spikes on four or more awake recordings had most or all seizures ipsilateral to most spikes, including 79 of 80 (99%) of those with > or = 3 side vs side spike ratios. Among the 120 patients with high (> or = 3) side vs side spike ratios, most or all seizures of 118 (98%) originated ipsilateral to most spikes. Predominant seizure origin also correlated with lateralized arrhythmic delta--from 90% ipsilateral seizures of those with one EEG with delta to 100% with > or = 4 such EEGs. Data from these two studies using opposite directions of analysis (seizures-->spikes/delta and spikes/delta-->seizures) demonstrate high correlations between laterality of interictal and ictal entities, particularly if temporal spikes clearly predominate on one side and if unilateral temporal delta activity persists over several recordings. Such correlations suggest that the awake interictal scalp EEG cannot be ignored when assessing laterality of temporal epileptogenesis.

Morphology of spikes in spike-and-wave complexes.

The morphology of spikes of spike-and-wave complexes from 9 patients with primary generalised epilepsy was studied from the scalp recordings of electrodes placed over their fields of maximum expression. Using a 4 x 4 grid electrode placement, selected epochs of the EEG were digitised and stored by a computer program which determined the times of the troughs and peak of the spike as well as its slope, amplitude and duration. Morphological parameters corresponding to those studied for focal spikes in other works gave similar values suggesting a close resemblance between spikes of generalised spike-and-wave complexes and focal spikes. In this study, second half-wave duration exceeded first half-wave duration and amplitudes of the second half-waves were either equal to or greater than those of the first half-waves. Our measures of slope resolved an apparent discrepancy among earlier studies: the entire first half-wave slope was usually equal to or greater than the second half-wave slope while the slopes of the upper half of the spike were usually equal or greater for the second half-wave than for the first. The first half-wave: second half-wave slope ratio was greatest at electrode positions with the maximum spike amplitude (the midline frontal electrodes), and therefore more closely resembled electrocorticogram spikes at these positions.

Topographical evolution of spike-wave complexes.

Computer-generated 3-dimensional field potential maps of spike-wave complexes from two 4 X 4 electrode grids on the scalp were studied. A visual analysis of these field maps throughout the spike-wave evolution permitted quantification of the spike, trough and slow wave components in terms of distribution, origin and propagation. In addition, a more objective morphological analyzer also quantified the discriptive parameters of distribution, origin and propagation for the spike component of the patients' spike-wave complexes. We found that field distributions of spikes differed from that of waves. Succeeding positive troughs evolved more symmetrically than did spikes but less than ensuing negative waves. Negative waves were more diffuse, more symmetrical in evolution, and more posteriorly centred than either spikes or troughs. Unlike the troughs and slow waves whose fields tended to remain stationary during their evolution, spikes always moved from their points of origin. Spikes originated at the most lateral points of the grids and propagated laterally and anteriorly in one of two ways: a simultaneous origin at both lateral positions, then propagation toward the midline, and then usually anteriorly, a clearly unilateral origin with spread contralaterally to the homologous electrode position of the contralateral hemisphere followed again by anterior propagation. Interhemispheric lag times of spikes ranged from 0 to 25 ms with an average of 10.5 ms. Inter- and intrapatient variability was considerable. This type of analysis reveals properties of spike-wave complexes which may not be appreciated by standard paper writeout.

Technique to display topographical evolution of EEG events.

We have developed a system to generate from 16 channels of the EEG 3-dimensional sequential topographical displays that facilitate the analysis of the propagation and field distribution of paroxysmal discharges. The ease and flexibility of data storage and analysis is as important as the concept of representing the EEG field potentials in a comprehensible fashion.

EEG morphology of partial epileptic seizures.

We studied the EEG features of partial seizures in 66 patients. An EEG evolution (morphology and/or frequency change) characterised 79% of attacks: 92% of 48 events with clinical features, but only 44% of 18 subclinical seizures. Aside from attenuation which initiated 7 seizures (11%), 31 (47%) began with sinusoidal waves, 25 (39%) with repetitive epileptiform potentials, and 10 (15%) with both phenomena. Metamorphosis between these forms occurred in about 1/3 of seizures beginning with either form alone. A wave form change occurred in 44% of clinical attacks but only 7% of subclinical seizures. Spikes and sharp waves were the most common repetitive epileptiform potentials encountered. Repetition rate of phenomena at onset usually lay in the theta or delta range except for a few with high frequency sinusoidal waves. Further ictal EEG progression was more likely to occur if a mixed frequency change or frequency increase characterised early evolution.