Multimodal Imaging of Spike Propagation: A Technical Case Report

Case Report/Technique

The patient was an 11-year-old boy with medically intractable epilepsy. Seizure onset was at age 4. His seizure semiology suggested that the right sensorimotor cortex for the left upper extremity was involved during ictal activity. Interictal EEG showed midline frontocentral spikes that were dominant on the right electrodes. A right superior frontal gyrus cortical dysplasia was identified on MR imaging (Fig 1A).

• Download figure
• Open in new tab
• Download powerpoint

Fig 1.

A, MR image shows a right superior frontal gyrus lesion (arrow). B, Active sites that show frequent spikes on EEG and MEG. Red circles and blue squares show EEG electrodes and MEG sensors, respectively. Note that the active MEG electrodes are located more inferiorly than the active EEG electrodes. C, The location of grids is shown on the cortical surface. D, A typical spike recorded on simultaneous EEG and MEG. The labels of the MEG channels correspond to the sensors shown on B, Each site has 2 gradiometers. E, Source waveforms of an EEG/MEG spike and a typical spike on IEEG. Source waveforms are extracted at sites 1 and 2 (superior frontal) for EEG and 3 and 4 (inferior frontal) for MEG. The superior frontal peak of EEG waveforms precedes the inferior frontal peak of MEG waveforms by approximately 20 ms. A similar time difference is seen on the IEEG spike. F, A source distribution map of a typical spike appearing on both EEG and MEG. Cortical activation is shown with red/yellow for EEG and blue/dark blue for MEG. The EEG activation appears in the right superior frontal area at the early latency (0 ms), whereas the MEG source shows later activation in the inferior frontal area (14 ms). White and red circles show the location of active IEEG electrodes in the superior (sites 1 and 2) and inferior (sites 3 and 4) frontal areas, respectively. G, The tractography shows a fiber connection between the right superior and inferior frontal gyri.

MEG was recorded with a 306-channel whole-head MEG system (Vectorview, Elekta Neuromag, Helsinki, Finland). EEG was simultaneously recorded with a 70-channel electrode cap, based on the 10–10 electrode-placement system. The sampling rate was 600 Hz, and the data were digitally filtered at 1–40 Hz for analysis. High-resolution 3T MR imaging data were acquired by using a 32-channel head coil with MPRAGE (TE = 3.37 ms, TR = 2000 ms, flip angle = 9°, voxel size = 1 × 1 × 1 mm) and multiecho FLASH pulse sequences (TE = 1.89 ms, TR = 20 ms, flip angle = 5°, voxel size = 1 × 1 × 1 mm) (Magnetom Trio; Siemens, Erlangen, Germany). We also acquired DTI sequences (60 gradient directions, TE = 82 ms, TR = 7850 ms, flip angle = 90°, section thickness = 2 mm, voxel size = 2 × 2 × 2 mm, diffusion sensitivity of b = 700 s/mm). IEEG was recorded by using grids and strips in the right medial and lateral frontal surfaces, covering the right superior and inferior frontal gyri (Fig 1C; XLTEK EEG system; Excel-Tech, Oakville, Canada). No depth electrodes were used. The sampling rate was 512 Hz, and the data were low-pass filtered at 30 Hz for analysis. The patient underwent CT after implantation of intracranial electrodes to confirm their location (voxel size = 0.5 × 0.5 × 0.5 mm).

We analyzed 50 spikes appearing simultaneously on MEG and EEG. Spatiotemporal source distribution was calculated by using a minimum norm estimate.⁴ We obtained a 3-layer head model from MPRAGE and FLASH MR imaging sequences by a boundary elemental method.⁵ The inverse solution was calculated from the forward solution, which models the signal-intensity pattern generated by a unit dipole distributed on the MR imaging–derived cortical surface.^6,7 The activation at each location was mapped at each time point of the spikes by using this inverse solution. For IEEG, we determined 2 active frequently spiking intracranial electrodes in each area of the superior and inferior frontal gyri. The largest time difference between the superior and inferior peaks at these electrodes was obtained from 50 spikes.

We coregistered the CT images onto the cortical surface and determined the location of the sites corresponding to the active intracranial electrodes (Fig 1C). Source waveforms of each MEG spike were extracted from the spatiotemporal source distribution at the sites of inferior frontal gyrus active electrodes. EEG source waveforms at the sites of the superior frontal gyrus electrodes were extracted in the same manner. We obtained the time differences of the peaks of MEG and EEG source waveforms for each spike.

For fiber tracking, we performed a tractographic analysis of DTI data by using TrackVis software (Wang and Wedeen, www.trackvis.org). This algorithm uses the standard fiber assignment by the continuous tracking method⁸ and calculates all possible fiber connections in the whole brain. Next we put an anatomic region of interest in the dorsal part of the lesion and the contralateral region (Fig 1G). The tractographic map was coregistered onto the MPRAGE images.