Abstracts

Rationale: Reading epilepsy is a rare type of reflex epilepsy in which seizures are provoked by reading. In a previous study, we investigated a 42 year-old male patient with reading epilepsy using clinical assessments and continuous video-electroencephalography (EEG). Results showed that reading-induced left parasagittal spike frequency significantly increased with involvement of the phonological reading pathway. In the present study, functional near infrared spectroscopy (fNIRS), functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) were used to localize the epileptogenic zone and analyze brain activity involved in reading.Methods: The participant read irregular words and non-words presented in a block-design paradigm during EEG-fNIRS, EEG-fMRI and EEG-MEG recordings. Articulation was overt during fNIRS and covert during MEG and fMRI recordings. EEG-fNIRS recordings were obtained using 19 EEG in-house electrodes placed according to the 10-20 system combined with 16 NIRS detectors and 55 sources covering the frontal, temporal, parietal and occipital lobes bilaterally. Haemoglobin concentration variations were calculated using the Modified Beer Lambert Law. fMRI acquisitions (399 volumes) were performed with a 3T Philips Achieva system using a standard head coil. Data were processed using SPM 8. Contrasts of interest were as follows for reading activation in fMRI and fNIRS: irregular word and non-word reading conditions versus baseline condition (fixation of a cross presented on a screen in a central position). MEG was acquired with a CTF-VSM whole-head 271-sensor MEG system; equivalent current dipole analysis was conducted for spike source localization.Results: Spike analyses revealed an epileptic focus in the left premotor and motor regions in fNIRS, fMRI and MEG recordings. There was no difference in spike localization between irregular word and non-word reading-induced spikes. Significant patterns of activation (i.e. fMRI positive BOLD or fNIRS HbR decrease) compared to baseline during irregular word and non-word reading were shown bilaterally in the occipital lobes and inferior frontal lobes (t ? 3.7, p < .05, uncorrected) and in the right temporo-occipital junction (t ? 3, p < .05, uncorrected). A deactivation pattern was revealed in the left anterior temporal lobe (t ? -3, p < .05, uncorrected) when the patient read irregular words and in the left supramarginalis gyrus and right posterior temporal lobe (t ? -2.5, p < .05, uncorrected) when he read non-words.Conclusions: The localization of epileptic focus in the left pre-central cortex was confirmed across three neuroimaging techniques. Significant activation was found in regions known to play a role in irregular word and non-word reading such as the occipital lobes, the inferior frontal lobes and the temporo-occipital junctions, confirming previous imaging studies of reading tasks in MEG and fMRI. Further investigation is needed to evaluate the significance of observed deactivation areas.