Abstracts

Rationale: Synthetic aperture magnetometry (SAM) is an adaptive spatial filtering algorithm for magnetoencephalography (MEG). SAM kurtosis (SAM(g2)) provides source locations of intracranial discharges with excess kurtosis value (steepness). Single dipole model (SDM) is a standard technique of MEG for estimating equivalent current dipole (ECD) of interictal spikes. SAM(g2) results have been reported concordant with the single clustered ECDs. Methods: We analyzed SAM(g2) and SDM in 39 patients with intractable neocortical epilepsy (age ranging 1-18 years; mean 9 years). We used whole head gradiometer Omega system (151 channels, VSM MedTech Ltd., Port Coquitlam, BC, Canada) with simultaneous EEG recorded from 19 electrodes. We recorded at least 15 times of 2-minute-dataset. SAM(g2) calculated kurtosis value (band pass filter, 20-70 Hz) at each voxel (5 mm distance) in the entire brain. In each dataset, we selected active voxels with local peak kurtosis higher than half of maximum value in each data set. We overlaid all selected active voxels from 5 datasets without artifacts on patient’s MRI. Using clustering analysis, we localized “cluster” which contains more than 6 ECDs within 1.7 cm distance between ECDs. We defined: “concordant case” when more than 50 % of active voxels overlapped with clustered ECDs; “partially concordant case” when less than 50 %; “discordant” when there was no overlap. Results: : Single clustered ECDs were found in 20 patients and multiple in 19 patients. Twenty-seven patients had lesions on MRI. Locations of active voxels overlapped with clustered ECDs between 8.7-100% (mean, 58.2%). Twenty-six (67%) patients were concordant (single cluster, 10 [lesion, 9; non-lesion, 1]; multiple clusters, 16 [lesions, 11; non-lesions 5]). Thirteen patients (34%) were partially concordant (single cluster, 10; multiple clusters, 3 and, lesions, 7; non-lesions 6). There was no discordant case. In 12 cases with both multiple clustered ECDs and lesions, the highest concordant achieved in 11 patients (92%). In 7 cases with multiple clusters without lesion, 5 patients (71%) were concordant. In 15 cases with a single cluster with lesion, 9 patients were concordant (60%). In 5 cases with both single clustered ECDs and non-lesion the lowest concordant resulted in only one patient (20%). Conclusions: This systematic SAM(g2) analysis succeeded in localizing epileptic sources correlating with both single and multiple clustered ECDs in children with intractable epilepsy. SAM(g2) showed highest concordant with the multiple clustered ECDs secondary to lesion. It is suggested that SAM(g2) could support or replace SDM in subset of epilepsy cases.