ABSTRACT
PURPOSE: To prove the clinical usefulness of SISCOM and compare SISCOM images derived from single- and dual-headed single-photon computed tomography (SPECT) cameras for localization of partial epileptic seizures. METHODS: We retrospectively studied 38 partial epilepsy patients, using subtraction SPECT coregistered to magnetic resonance imaging (MRI; SISCOM). SPECT imaging of the first 15 patients was performed by single-headed camera, and the next 23 patients by dual-headed camera. Side-by-side ictal-interictal SPECT evaluation and SISCOM images were blindly reviewed and classified as either localizing to one of 16 sites or nonlocalizing. A third reviewer evaluated cases of disagreement between primary reviewers. Results were compared with seizure localization by any of the following three traditional techniques: surgical outcome, invasive, and noninvasive video-EEG monitoring. The results from the single- and dual-headed SPECT cameras were compared. RESULTS: Reviewers localized areas of hyperperfusion with SISCOM images more often than with side-by-side SPECT evaluation (71.0 vs. 47.4%; p = 0.01). When we compared results of SPECT evaluation with traditional techniques, SISCOM showed greater concordance than side-by-side SPECT evaluation (60.53 vs. 36.84%; p = 0.006). There were no differences in localization between images derived from single- and dual-headed cameras. Concordance of seizure localization, compared with traditional techniques, also was not different between these groups [kappa = 0.38, 95% confidence interval (CI), 0.18-0.58] vs. kappa = 0.63, 95% CI (0.45-0.81)]. CONCLUSIONS: SISCOM is a worthwhile technique for preoperative evaluation in partial epilepsy patients and improves the sensitivity and specificity of seizure localization of SPECT images derived from both single- and dual-headed SPECT cameras.
