Abstracts

Rationale: Localization of cortical function is often performed prior to excision of mass lesions adjacent to eloquent cortex or surgery for intractable epilepsy. This localization is typically accomplished by electrical stimulation through subdural electrodes or by direct stimulation with devices such as an Ojemann stimulator. These stimulation-based mapping methods have several limitations. They are time-consuming, have limited resolution, may give misleading results if the stimulation spreads, and can be associated with afterdischarges that can lead to seizures. Recent studies have shown that offline analyses using passive recordings of electrocorticographic activity (ECoG) can detect task-related signal changes (Crone et al., 1998; Crone et al., 2001; Sinai et al., 2005; Miller et al., 2007; Leuthardt et al., 2007). While the exact relationship between ECoG-based mapping and stimulation-based mapping is not yet clear, it is likely that passive mapping may supplement or even replace traditional stimulation-based mapping methodologies. The ability to detect in real time the ECoG signal changes associated with task performance or with abnormal events such as seizures could be an important new addition to the localization methodologies available to epilepsy surgery. We describe here a novel signal processing and visualization method called SIGFRIED that uses passively recorded ECoG activity to detect and map task-related (e.g., language and sensorimotor) function or abnormal (e.g., epileptic) activity in real time.Methods: Using mathematical modeling of several mu/beta and/or gamma-rhythm frequency features, the properties of resting ECoG activity are estimated. In real-time operation, SIGFRIED determines the difference of current ECoG activity to this resting signal model at each location. These difference measures are shown on a topography that consists of circles at the electrode locations. The radius of each circle is proportional to the signal change so that changes can be quickly identified and localized. Results: In three patients with subdural grid recordings, activity at distinct and reproducible grid locations corresponded to language and motor tasks. The acquisition of interpretable averaged topographic findings took 30 seconds to three minutes. The location of the identified signal changes was consistent with the results reported in recent offline ECoG mapping studies (Miller et al., 2007, Leuthardt et al., 2007) and in general agreement with those achieved using stimulation-based mapping.Conclusions: SIGFRIED can visualize brain signal changes in real time, and can thus be used in an exploratory fashion to identify and localize brain function associated with different tasks. These results are in agreement with those achieved using stimulation-based mapping. Our initial tests of this method thus encourage further studies on the efficacy of SIGFRIED for mapping purposes and on the detailed relationship to stimulation-based mapping. Furthermore, preliminary results provide encouraging evidence of the utility of this real-time technique for the detection and topographical mapping of epileptic activity.