Ictal single-photon emission computed tomography imaging in extratemporal lobe epilepsy using statistical parametric mapping

Purpose. To examine the application of statistical parametric mapping (SPM) to analyze ictal single-photon emission computed tomography (SPECT) scans in surgical candidates with extratemporal lobe epilepsy. Methods. The authors selected patients who underwent successful ictal SPECT acquisition in the process of surgical treatment of intractable partial epilepsy. Thirteen patients were identified who met inclusion criteria for confident seizure localization from either intracranial electroencephalograrn recordings or epilepsy surgery outcome. In these cases, ictal scans were registered to an in-house-developed normal SPECTatlas composed of 14 spatially normalized brains of normal subjects. SPM96 was used to test on a voxel-by-voxel basis for statistically significant increases in blood flow associated with each patient's ictal scan. The results were then mapped back onto the patient's magnetic resonance image (MRI) for final interpretation. Statistical parametric mapping (SPM) analysis of ictal SPECT scans was compared to both conventional visual interpretation and the analysis of subtraction ictal SPECT co-registered to MRI (SISCOM). Results, Ten of 13 patient scans showed localizing focal ictal increases in regional cerebral blood flow, all of which were concordant with ultimate epilepsy localization. Of the 3 cases not localized with SPM, 1 was localized by conventional visual interpretation and another, not localized by visual interpretation, was correctly localized with SISCOM. Two cases not localized by SISCOM were localized by both visual and SPM analysis. Conclusions. This work provides supportive evidence for proof of principle thatSPIM can be used to provide objective, accurate analysis of ictal SPECT scans in patients with extratemporal lobe epilepsy.