SENSITIVITY AND SPECIFICITY OF EEG SOURCE IMAGING FOR FOCUS LOCALIZATION IN THE PRESURGICAL EVALUATION
Abstract number :
3.266
Submission category :
5. Human Imaging
Year :
2009
Submission ID :
10352
Source :
www.aesnet.org
Presentation date :
12/4/2009 12:00:00 AM
Published date :
Aug 26, 2009, 08:12 AM
Authors :
Laurent Spinelli, V. Brodbeck, G. Lantz, C. Pollo, K. Schaller, M. Vargas, M. Wismmeyer, C. Michel and M. Seeck
Rationale: Even though EEG Source Imaging (ESI) is a powerful and easy accessible tool for focus localization in the presurgical work up partial epilepsy patients, it is used in few centres only. Here we determine the sensitivity and specificity of ESI in a large group of operated focal epilepsy patients and compare the results to those of most often used non-invasive methods for focus localization. Methods: Of all therapy-refractory focal epilepsy patients that had undergone presurgical invasive evaluation from 1995 to 2008 in our department, a total of 36 operated patients was identified. For these 36, we demonstrate the presurgical results of ESI, MRI, FDG-PET and ictal SPECT and define for each method whether it localized the focus exclusively within the subsequently resected area. ESI was carried out on the basis of 32- to 256 surface EEG recordings The results are grouped according to the post-operative outcome into good (Engel Class I-II, group A) and unfavourable outcome (Engel Class III-IV, group B). Sensitivity is defined as percent of cases group (A) where the method localized within the resection zone. Specificity is defined as percent of cases where the method localized outside or not exclusively within the resection zone and the resection did not lead to worthwhile outcome. Results: Overall, 27 patients had a good (group A), 9 an unfavourable outcome (group B). Sensitivity for unambiguous focus localization was highest for ESI (67%) followed by PET (63%) and ictal SPECT (50%). MRI had a sensitivity of 48%. The specificity was 87% for ESI as opposed to 67% for MRI, PET and ictal SPECT. “False” localizations within the resection zone in patients of group B occurred in 13% for ESI, but 33% for PET and ictal SPECT. MRI showed 22% false positive cases. Conclusions: Our results highlight the value of ESI for focus localization in epilepsy patients. Sensitivity and specificity of ESI are high and comparable or better than those obtained with frequently used non-invasive imaging techniques as PET and SPECT.
Neuroimaging