Ictal localization of temporal lobe seizures with scalp/sphenoidal recordings

Abstract

We assessed the reliability and accuracy of scalp/sphenodial recordings for ictal localization by retrospectively analyzing 706 noninvasive ictal recordings from 110 patients who subsequently underwent stereoencephalographic evaluation. Strictly defined unilateral temporal/sphenoidal ictal patterns correctly predicted findings of depth electrode examination in 82 to 94% of cases. These strictly defined predictive patterns could be detected with excellent interrater reliability. The patterns are misleading in only a minority of cases, but cannot be used in isolation for definite ictal localization.

Temporal and spatial determination of EEG-seizure onset in the frequency domain

“…All patients had a similar evaluation that included interictal and ictal video-EEG, neuropsychological testing, interictal and ictal SPECT (single photon emission computer tomography), brain magnetic resonance imaging (high resolution in seven cases), in seven cases interictal PET (positron emission tomography) and neuropathological analysis of the resected tissue. Patients ful®lling the following two criteria were included in the analysis: (1) patients with maximal ictal activity in the sphenoidal electrode and/ or temporal electrode with clear focal rhythmic activity (Risinger et al, 1989); (2) patients evaluated with highresolution MRI (n 7) who had signs of unilateral hippocampal sclerosis on MRI scans (diminished volume of the hippocampus and/or T2-signal enhancement) or patients a An asterisk (*) in the MRI column indicates that low-resolution MRI has been carried out. In patient 9 a discrepancy can be noted between SPECT ®ndings (left hemispheric) and other results from the presurgical workup (right temporal).…”

“…1; ESO focal or non-focal sharp-and slow waves, spikes or spike wave complexes) and the onset of clear continuous focal rhythmic activity ( Fig. 1; RSO well modulated rhythms of temporal location (Risinger et al, 1989;Ebersole and Pacia, 1996) were determined by visual video-EEG analysis. After excluding the signals recorded with the sphenoidal electrodes, the data were further analyzed using the following methodological approach.…”

“…We did not include the onset of ictal behavior to select the ictal time period, since it has been shown in previous studies that ictal EEG analysis alone can reliably de®ne the seizure onset (Risinger et al, 1989;Ebersole and Pacia, 1996;Lantz et al, 1999). The selected 46 s time period was divided into 23 non-overlapping 2 s epochs.…”

“…2, Table 2; Risinger et al, 1989;Ebersole and Pacia, 1996;Lantz et al, 1999). Over all patients the mean of this frequency was determined at 7.0^0.8 Hz).…”

“…The most frequent pattern is characterized by a widespread attenuation of background activity, without topographic value, followed by the appearance of rhythmic frequencies containing localizing information (Quesney et al, 1993). When analyzing these lateralized rhythmic patterns, Risinger et al (1989) found a high correlation between ictal patterns of 5 Hz or faster frequencies in a sphenoidal and/or temporal location and the presence of an ipsilateral mesial temporal seizure onset. Although this pattern was quite speci®c, it unfortunately had a poor sensitivity of only 40±50%.…”