Fig 1.
Examples of two depth electrode-recorded ictal EEG onset patterns from two different patients diagnosed with unilateral mesial temporal lobe epilepsy.
(A) Depth EEG recording displayed in a bipolar montage of a focal hypersynchronous ictal onset pattern consisting of low frequency (<2 Hz), high amplitude spikes in RPG1-2 that precedes the spread to RAH1-2 and REC1-2. This seizure spread to LPG1-2 and LAH1-2 1 min 21 sec after ictal onset (not shown). Total seizure duration was 2 min 56 sec. (B) Depth EEG recording of a low voltage fast ictal onset pattern, which compared to preceding EEG baseline, begins with low amplitude, high frequency (~33 Hz) activity nearly simultaneously in LEC1-2 and LA1-2. Note ictal activity appears in LOF3-4 and 5–6 within 5 sec and in RA1-2 11 sec after ictal onset. Total seizure duration was 1 min 10 sec. Depth electrodes labels and anatomical location as follows: R (right)/L (left) A (amygdala), AC (anterior cingulate), AH (anterior hippocampus), EC (entorhinal cortex), OF (orbitofrontal cortex), PG (posterior parahippocampal gyrus), STG (superior temporal gyrus), and TPO (temporal-parietal-occipital junction); numbers refer to electrode contacts 1 (distal, medial surface) to 6 (proximal, lateral surface).
Fig 2.
Radar plot illustrating the results of cluster analysis using anatomical /electrode sites(s) of the SOZ and initial seizure spread.
Involvement of each recording site (denoted by radii on each plot) in the SOZ (top row A-C), sites of initial ipsilateral spread (middle row D-F) and initial contralateral spread (bottom row G-I) is expressed as the proportion of total seizures in each cluster (n) indicated by the concentric circles in each plot. Right and left halves of each plot depict medial and lateral aspects respectively of temporal and frontal lobe that included orbitofrontal (OF), amygdala (Am), entorhinal cortex (EC), anterior (AH), middle (MH) and posterior hippocampus (PH), parahippocampal gyrus (PHG), and anterior (a), middle (m), and posterior (p) regions of inferior (ITG) and middle temporal gyri (MTG).
Table 1.
Clinical and seizure statistics.
Table 2.
Extent of SOZ and initial seizure spread.
Fig 3.
Color-coded probability or P-maps depicting the spatial pattern of significantly reduced gray matter thickness in patients with respect to controls.
Lateral and medial views of cerebral hemisphere ipsilateral (A & C) and contralateral (B & D) to the SOZ in patients with hypersynchronous (HYP, top row) versus patients with low voltage fast (LVF, bottom row) onset seizures. P-maps in bottom row are the same as in rows above, but reoriented to more clearly show gray matter thickness changes on ventral, dorsal, and anterior aspects. In upper right, P-values scaled such that areas shaded green correspond to P<0.01 and blue P>0.05 (not significant).
Fig 4.
Color-coded P-maps of the lateral and medial surface of the cerebral hemisphere ipsilateral (A) and contralateral (B) to the SOZ depicting areas with a significant reduction in cortical thickness in relation to depth EEG seizure onset pattern and epilepsy duration after controlling for age and gender.
Top row illustrates sites of significantly reduced gray matter thickness per year of epilepsy in patients with HYP onset seizures compared to those with LVF onset seizures, whereas bottom row reflects areas with reduced thickness in patients with LVF onset seizures with respect to those with HYP onset seizures. Permutation test correcting for multiple comparisons was significant (ipsilateral: p = 0.037; contralateral: p = 0.036). Areas masked in white (i.e., corpus callosum, diencephalon) were not included in this analysis. Color-coded p-value scale in upper right corner of the figure.
Fig 5.
Beta maps illustrating the magnitude of the interaction between ictal EEG onset pattern and epilepsy duration on cortical gray matter thickness after controlling for age and gender.
Lateral (left column) and medial surfaces (right column) of cerebral hemisphere ipsilateral (top row) and contralateral to the SOZ (bottom row). Cortical areas colored yellow, orange, and red indicate thinner GM (in mm) per year of epilepsy in patients with HYP onset seizures, whereas areas colored blue, indigo, and violet correspond to reduced GM per year of epilepsy in patients with LVF onset seizures. Shades of green depict areas of minimal or no thinning per year of disease (color-coded scale upper right corner). Permutation tests correcting for multiple comparisons were significant (ipsilateral: p = 0.037, contralateral: p = 0.037). Areas masked in white are brain areas not applicable for this analysis.