Abstracts

Experimental Tests of EEG Localization Accuracy

Abstract number : 1.006
Submission category :
Year : 2000
Submission ID : 2335
Source : www.aesnet.org
Presentation date : 12/2/2000 12:00:00 AM
Published date : Dec 1, 2000, 06:00 AM

Authors :
B Neil Cuffin, Donald L Schomer, John R Ives, Howard Blume, Beth Israel Deaconess Medical Ctr, Boston, MA.

RATIONALE: EEGs can be used to calculate the locations of electrical sources in the brain. However, the accuracy of these calculations is not well known because it is usually not possible to compare calculated locations with actual locations since actual source locations are generally not well known. METHODS:: In this study, sources at known locations are produced by injecting current into implanted electrodes in the brains of patients undergoing epilepsy presugical evaluation; the locations of the electrodes are accurately determined from CTs. The EEGs produced by these sources are used to calculate the locations of the sources in spherical models of the head containing three layers representing the brain, skull, and scalp. The skull and scalp layers are 6 and 5 mm thick, respectively. The electrical conductivties of the brain and scalp regions are equal, while three different skull conductivity ratios of 1/80, 1/40, and 1/20-th of the brain and scalp regions are used. EEGs have been obtained and localization errors determined for 177 sources in thirteen subjects. RESULTS: The smallest average localization error of 10.6 mm for all 177 sources is obtained for a skull conductivity ratio of 1/40. The average errors for the 1/80 and 1/20 ratios are 12.1 and 12.2 mm, respectively. The average localization error for 108 sources at superior locations in the brain, i.e., above the center of the brain region, is 9.2 mm. The average error for 69 inferior location sources is 12.6 mm. CONCLUSIONS: These results indicate that for these spherical model dimensions, the best skull conducutivity ratio is 1/40 rather than the generally used ratio of 1/80. However, the results also indicate that localization accuracy is not significantly different over a wide range of skull conductivity ratios around the the 1/40 ratio. The localization errors obtained here are only a few millimeters smaller than those obtained in the best previous study. This indicates that an average localization accuracy of approximately 10 mm is the best that can be achieved using spherical head models. More realistic head models will be required for greater localization accuracy and these results indicate that this need is greatest for inferior location sources.