Abstracts

Rationale: Functional cortical mapping (FCM) is an essential part of epilepsy surgery programs. The gold standard for FCM remains electrical cortical stimulation mapping (ECSM) to inhibit selected functional processes. The method has some limitations: eg, risk of producing seizures and perhaps incomplete mapping of cortex that generates responses. New technologies that analyse changes in spontaneous cortical activation during tasks, such as brain computer interface (BCI), may provide real-time localization of cortical function with minimal additional risk compared with ECSM. Recently, SIGFRIED software (Schalk et al. 2008) was developed for physiologic real-time functional mapping in epilepsy surgery candidates. Results of such mapping in adults highly correlated with ECSM produced results (Brunner et al. 2009). Whether BCI can be applied for pediatric epilepsy surgery patients remains unknown. We assessed feasibility and methodological advantages of real-time functional cortical mapping with BCI2000-based SIGFRIED software in pediatric epilepsy surgery patients. Methods: Two adolescents with intractable epilepsy participated in the study. Each had electrocorticographic (ECoG) electrode grids or strips placed subdurally for seizure localization prior to epilepsy surgery. The electrode locations were based solely on the requirements of the clinical evaluation without any consideration of this study. Grid locations for each patient are shown in Fig. 1 (Patient A: Fig. 1A; Patient B: Fig. 1D). During a single hour session we tested both active and passive conditions, including listening to different sounds (e.g., classical music, stories, and white noise), reading sentences, and naming objects.Results: The procedure was well tolerated by both patients; no seizures were induced or occurred spontaneously. Most stimuli elicited some degree of cortical activation (Fig. 1). Patient A showed activation in the primary auditory cortex and posterior superior temporal gyrus and during vocal naming tasks (Fig. 1B), while silent naming tasks produced similar posterior activation patterns without auditory cortical activation (Fig. 1C). In Patient B, classical music, listening to stories and white noise elicited similar patterns of activity in the lateral and basal temporal regions (Figs. G, H, and I, respectively).Conclusions: Pediatric patients are an ideal population for real-time functional mapping using ECoG signals. The techniques shown are adaptable to a wide range of age groups and intellectual capacities, and do not have the risk of inducing seizures. Furthermore, it provides the ability to measure the true physiological function of the underlying cortex, while ECSM infers functional map from the observed functional changes from the stimulation of brain. Further testing is needed to develop suitable paradigms, but these results demonstrate the utility of these BCI-based techniques. References: Brunner P et al (2009). Epilepsy Behav 15: 278-86 Schalk G et al (2008). Neuroimage 43: 245-9 Important notice: M. Korostenskaja and A. Wilson contributed equally to the present study