Abstracts

Rationale: Evaluation of language lateralization is highly important for planning epilepsy surgery. Magnetoencephalography (MEG) is one of the potential tools for this purpose, however, conventional analysis by using a single dipole method may be sometimes erroneous. We assessed the usefulness of dynamic statistical parametric maps (dSPMs) for investigating language lateralization in patients with epilepsy. Methods: MEG was performed for presurgical language mapping in ten epilepsy patients (male:3, female:7, age:9-34). Informed consent was obtained from each patient or guardian. All patients showed left language predominance in Wada test. MEG was recorded with a 306-channel whole-head system at a sampling rate of 600Hz. In all patients, high-resolution 3T anatomical MRI data were acquired with magnetization-prepared rapid acquisition gradient-echo (MPRAGE). For the paradigm of language testing, we performed a semantic language task. Patients were asked to decide whether each word was abstract or concrete in a visually presented series of words. The MEG data was averaged by using the timing of word presentation as a trigger. We calculated dSPMs by applying a distributed source model mapped on the MRI-derived cortical surface. The activation between 250ms and 550ms after the trigger was analyzed. In this study, we focused on the receptive language area (Wernicke s area), and included superior temporal, middle temporal, supramarginal and inferior parietal cortices on both hemispheres for further analysis. These areas were determined on the cortical surface reconstructed from each patient s MRI. Laterality index of dSPMs (LI-dSPMs) was obtained by LI = (L-R)/(L R), where L and R is the number of unit dipoles with an F value higher than the threshold value in these cortical areas of left and right hemisphere, respectively. We also calculated equivalent current dipoles (ECDs) sequentially between 250ms and 550ms based on a single dipole model by using the left and right hemispheric sensors separately. Laterality index of ECDs (LI-ECDs) was obtained by LI = (L-R)/(L R), where L and R is the number of ECDs with goodness of fit higher than 60%. (Foxe D, et al, AES annual meeting, 2003). For each patient, language predominance was determined based on the laterality index as follows; ?0.1:left, 0.1>LI>-0.1:bilateral, ?-0.1:right. We compared the results of LI-dSPMs and LI-ECDs with WADA test. Results: All patients showed left language lateralization in LI-dSPMs. LI-dSPMs of the patients ranged from 0.10 to 0.37. In LI-ECDs, five patients showed left predominance, and five patients were bilateral. LI-ECDs ranged from -0.09 to 0.31. No patient showed right-sided language predominance in both LI-dSPMs and LI-ECDs. Conclusions: In our patients, LI-dSPMs were more consistent with the language lateralization determined by WADA test than LI-ECDs. DSPMs may provide more reliable information in the analysis of language MEG data than a single dipole method.