[A comparative study on intraoperative electrocorticography with long-term intracranial EEG for localizing
Li-Xin Cai1, Yong-Jie Li, Guo-Jun Zhang
1Beijing Institute of Functional Neurosurgery, Xuanwu Hospital, Capital Medical University, Beijing 100053, China. cailx1010@hotmail.com
Objective:
To discuss the localizing methods for epileptogenic zone with intraoperative electrocorticography (ECoG) by comparing it with long-term intracranial EEG.
Methods:
Twenty-two patients was enrolled, who received epilepsy surgeries consecutively in our institute since April 2005 to August 2006. All patients underwent intracranial electrode implantations with long-term video-EEG monitoring before the resective surgeries and the postoperative pathologies were all proved to be focal cortical dysplasia. The extra-operative and ECoG digital EEGs were statistically analyzed and compared with each other, including interictal epileptic discharge (IED) frequency, amplitude and their correlations with ictal discharge (ID).
Results:
In the long-term intracranial EEG, significant difference was found between IED frequency derived from ID onset verus none-onset electrodes (Z = 12.213, P < 0.001), ROC analysis showed that IED frequency is a moderately good indicator for ID, the area under the ROC curve was 0. 758 (95% CI, 0.707-0.810). There was significant difference in IED amplitude between ID and none-ID electrodes (Z = 11.765, P < 0.001). The IED frequency was significantly lower during ECoG than during long-term extra-operative monitoring (rho = 0.518, P = 0. 014). A significant positive correlation in the IED frequency pattern was found between intra-operative ECoG and extra-operative recording when the average IED frequency was over 3 spikes /min. In ECoG, the correlation of IED with ID was similar in the 'frequent' (> or = 10 spikes/min) and 'occasional' (3-10 spikes/min) groups and very different in the 'rare' (< or = C3 spikes/min) group (rho = 0.408, P = 0.147).
Conclusions:
The IED from long-term EEG recording had moderate correlation in frequency and amplitude with ID position. IED frequency was usually decreased on ECoG when compared with long-term monitoring. However, ECoG could reliably reflect the IED pattern and the correlations between IED with ID in term of frequency and amplitude in the long-term EEG monitoring, when IED frequency exceeded 3 spikes/min. Under this circumstance, IED could be used for localizing the epileptogenic zone.
