Reproducibility of BOLD localization of interictal activity in patients with focal epilepsy: intrasession and intersession comparisons

Abstract

Simultaneous EEG-fMRI recordings allow the identification of haemodynamic changes induced by neuronal activity during ictal or interictal epileptiform events (IEDs). We evaluated the reproducibility of continuous EEG-fMRI (cEEG-fMRI) in patients with focal epilepsy. We studied 15 patients with focal epilepsy (8 cryptogenic and 7 symptomatic) and frequent interictal abnormalities. Each patient underwent two cEEG-fMRI acquisitions (runs) in the same day (session) and 8 patients repeated the examination after one month. cEEG-fMRI reproducibility was defined by the existence of partially overlapping clusters between activation maps obtained from different runs. We detected IEDs in 40 out of 46 EEG-fMRI runs and a related significant BOLD-response in all 40 runs. A prevalent positive BOLD response was detected in 12 patients and a prevalent negative response in 3 subjects. Statistical maps included a mean of 10 significant clusters. Nearly 30% of clusters were reproducible in both intrasession and intersession comparisons, with a mean overlap of 30%. Reproducibility did not differ between positive and negative BOLD-responses. Among the reproducible clusters, those with the highest percentage of overlap were concordant with the EEG electric field in all patients and they were localized in the same lobe as the brain lesion in patients with symptomatic epilepsy. We hypothesize that reproducible clusters could be more consistently related to the irritative zone than non-reproducible ones. The evaluation of cluster reproducibility could improve our knowledge of IED-related BOLD response. Moreover, it could enhance the reliability of cEEG-fMRI to identify the irritative zone in focal epileptic patients.