Abstracts

Rationale: Recently there have been reports of focal fMRI activation associated with interictal epileptiform discharges (`spikes') (1)(2). Such experiments require EEG equipment that is compatible with the MRI environment for safety and artefact reasons. Low signal change associated with `spikes' has meant that large numbers of spike-triggered acquisitions are needed to gain significance. Methods: We have adapted conventional EEG equipment allowing near continuous recordings inside the MRI at both 1.5 and 3 Tesla. Results: We have successfully imaged two patients at 1.5T and two at 3T, collecting on average 15-20 spikes per patient. Carbon fibre leads with non-metallic electrodes are attached to the scalp and a fibre optic link takes the EEG signal from the head box to recording computer. Careful electrode placement, modifying head restraint and ensuring low electrode impedance have reduced cardio-ballistic (pulse) artefact without filtering. Image acquisition uses a paused EPI-BOLD sequence with 1-3 repetitions triggered per spike, interspersed with similar baseline triggered acquisitions. After motion correction, global intensity normalisation and gaussian smoothing image analysis uses the student's t-test. In the patients studied, `active' pixels are associated with 0.5-1% signal change at 1.5T and approximately double this at 3T. Conclusion: Spike triggered fMRI is possible at 3T although cardio-ballistic artefact is more troublesome. The higher field strength enhances BOLD contrast, improving significance from fewer `spikes'. 1) Krakow K, Woermann FG, Symms MR, Allen PJ, Lemieux L, Barker GJ, Duncan JS, Fish DR. Brain 1999. 122, 1679-1688 2) Seeck M, Lazeyras F, Michel CM, Blanke O, Gericke CA, Ives J, Delavelle J, Golay X, Haenggeli CA, de Tribolet N, Landis T. Electroenceph and Clin Neurophys 1998. 106, 508-512.