Abstracts

Rationale: Functional cortical mapping for language has been indispensable to resective epilepsy surgery, brain tumor resection and continued investigations in neurocognitive disorders such as autism. It has helped us to better understand functional cortical organization and prevent unwanted deficits in patients requiring surgery. Traditionally, language mapping has been limited to lateralization procedures such as intracarotid anesthetic administration (Wada test) or invasive brain surgery with intracranial electrode placement and bedside or operating room testing. The latter is difficult to perform on young children or children with significant cognitive deficits. We present a brief case series using magnetoencephalography (MEG) on sleeping/sedated patients to map language areas of the brain. Methods: Three children (two females, one male) underwent passive language mapping using a 4-D, Magnes 3600, 248 channel magnetoencephalography system. Their ages range from 3 years to 16 years. The two female children have left temporal lobe epilepsy and the male has autism. A 180 word continuous recognition module (CRM) was played via air conducting tubes to buds in each ear equally. The CRM consisted of common English words that patients should be familiar with, presented as one word every two seconds. The volume is 85 dB. Each patient was allowed to fall asleep prior to the language mapping and sleep was confirmed in stages I and II via simultaneous 19 channel electroencephalogram (EEG) using the International 10-20 electrode placement system. Resultant magnetic field responses to the words were filtered and averaged. The average waveform was visually inspected and dipole sources were localized using an equivalent current dipole (ECD) model. Only sources between 250 and 800 ms post stimulus presentation were used to determine language laterality and Wernicke’s area. Subsequent evoked fields were co-registered onto thin cut (2 mm) structural MRIs representing magnetic source imaging (MSI). Results: Both epilepsy patients demonstrated a right superior temporal gyrus dominance for language. This was confirmed by electrocortical stimulation prior to epilepsy surgery. Resective epilepsy surgery was performed in both cases without language deficits. The autism patient demonstrated a bitemporal activation pattern with a slight dominance toward the left. Conclusions: Our small case series demonstrates the feasibility of passive language mapping. This is particularly valuable as a non-invasive presurgical evaluation tool, especially in patients who cannot cooperate with more traditional methods like the Wada test. Moreover, the accuracy of localization is far superior and has been consistent with direct cortical mapping. The bilateral activation pattern of an autistic patient may represent disorganized language processing networks, however, continued research is ongoing. We are excited by the possibilities of further understanding functional brain networks in young children with neurological disorders.