Original Research EEG-fMRI Using z-Shimming in Patients With Temporal Lobe Epilepsy Andrew P. Bagshaw, PhD, 1,2 * Leili Torab, PhD, 2 Eliane Kobayashi MD, PhD, 2 Colin Hawco, BSc, 2 Franc ¸ois Dubeau, MD, 2 G. Bruce Pike, PhD, 2 and Jean Gotman, PhD 2 Purpose: To use z-shimming, a technique that reduces signal loss due to susceptibility artifacts that can result in reduced or absent activation in electroencephalography (EEG) functional MRI (fMRI) sessions in patients with tem- poral lobe epilepsy (TLE), to determine whether it would result in an increased ability to detect significant regions of blood oxygenation level-dependent (BOLD) signal change. Materials and Methods: Eight patients with TL EEG spikes underwent an EEG-fMRI scanning session using z-shim- ming. The signal intensities in the z-shimmed images were compared with those in the standard images. BOLD acti- vation maps were created from the two sets of images using the timings of the spikes observed on the EEG. Results: The mean signal increase in the TLs as a result of z-shimming was 45.9%  4.5%. The percentage of TL voxels above a brain intensity threshold rose from 66.1%  7.6% to 77.6%  5.7%. This appreciable increase in signal did not lead to any significant differences in the statistical maps created with the two sets of functional images. Conclusion: The results suggest that loss of signal is not the limiting factor for the detection of spike-related BOLD signal changes in patients with TLE activity. Key Words: fMRI; EEG; z-shimming; epilepsy; localization J. Magn. Reson. Imaging 2006;24:1025–1032. © 2006 Wiley-Liss, Inc. TEMPORAL-LOBE EPILEPSY (TLE) is one of the most common forms of epilepsy, and is also one of the most amenable to surgical intervention (1,2). Although the combination of electroencephalography (EEG) and functional MRI (fMRI) has recently been applied to the study of epilepsy by a number of groups (3– 8), it has not yet shown itself to be a mature technique that can be of significant use in the epilepsy clinic. Given the possi- bility of successful surgical intervention, EEG-fMRI is potentially a very useful tool for preoperative evalua- tion, with the aim of localizing the epileptogenic region accurately and noninvasively. However, a consistent feature of the reported studies is that a relatively large proportion of scanned patients who demonstrated in- terictal activity during the scanning session did not show any significant fMRI responses. It has been shown that using a range of hemodynamic response functions (HRFs) in the fMRI analysis can result in an increased number of patients showing significant responses (10). This suggests that misspecification of the HRF ac- counts for at least some of the patients with no re- sponses. Previous studies have also suggested that sig- nal loss in patients with TLE may be a factor in preventing significant spike-related fMRI activations from being detected (5,9,10). Magnetic field inhomogeneities at the boundaries of tissues with very different magnetic susceptibilities lead to distortions and a loss of signal in fMRI acquisi- tions based on echo-planar imaging (EPI) (11,12). This problem is particularly severe in the inferior TLs and the orbitofrontal cortex (13). A number of strategies have been developed to reduce the signal loss (13–18), including z-shimming, which in its original implemen- tation refers to the two-dimensional acquisition of mul- tiple images with different preparation gradient pulses to correct for field inhomogeneities in the slice selection direction (19 –22). The purpose of the current work was to apply a two-dimensional z-shimming method to a group of patients with TLE in order to increase the signal in the TLs. The motivation was to determine whether this would increase the ability of EEG-fMRI to detect regions of significant spike-related BOLD signal change. MATERIALS AND METHODS Data Acquisition Eight patients with a clinical diagnosis of TLE under- went continuous EEG-fMRI monitoring using a z-shim- 1 School of Psychology, University of Birmingham, Birmingham, United Kingdom. 2 Montreal Neurological Institute, McGill University, Montre ´al, Canada. Contract grant sponsor: Canadian Institutes of Health Research; Con- tract grant number: MOP-38079; Contract grant sponsor: Savoy Foun- dation for Epilepsy; Contract grant sponsor: Preston Robb Fellowship, Montreal Neurological Institute. *Address reprint requests to: A.P.B., School of Psychology, University of Birmingham, Birmingham B15 2TT, UK. E-mail: a.p.bagshaw@bham.ac.uk Received February 23, 2006; Accepted July 28, 2006. DOI 10.1002/jmri.20744 Published online 11 October 2006 in Wiley InterScience (www. interscience.wiley.com). JOURNAL OF MAGNETIC RESONANCE IMAGING 24:1025–1032 (2006) © 2006 Wiley-Liss, Inc. 1025