STATISTICAL PARAMETRIC MAPPING BASED SPECT PROCESSING IN NONLESIONAL TEMPORAL LOBE EPILEPSY
Abstract number :
2.163
Submission category :
5. Neuro Imaging
Year :
2012
Submission ID :
16184
Source :
www.aesnet.org
Presentation date :
11/30/2012 12:00:00 AM
Published date :
Sep 6, 2012, 12:16 PM
Authors :
V. Sulc, D. Hanson, S. Stykel, B. Brinkmann, D. Jones, D. Holmes, R. Robb, M. Senjem, A. Larson, B. Mullan, D. Horinek, G. Cascino, E. So, G. Worrell
Rationale: Ictal SPECT has proven to be a useful tool in the presurgical evaluation of epilepsy surgery candidates, and subtraction of ictal and interictal SPECT with coregistration to MRI (SISCOM, O'Brien et al, 1998) has become a standard in seizure onset localization. The specificity of SISCOM, however, is likely reduced by the physiological variability of interictal perfusion. To overcome this issue, methods using statistical parametric mapping (SPM) applied to scans from a population of control subjects were developed. To evaluate the benefit of SPM based methods over traditional SISCOM in nonlesional temporal lobe epilepsy we compared Ictal-Interictal SPECT Analyzed by SPM (ISAS, McNally et al., 2005) and Statistical Ictal SPECT coregistered to MRI (STATISCOM, Kazemi et al., 2010) to SISCOM. Methods: We retrospectively identified 21 cases of nonlesional temporal lobe epilepsy that had a normal preoperative "seizure protocol" MRI, and underwent a standard anterior temporal lobectomy between January 1997 and December 2005 at Mayo Clinic, Rochester, MN. Interictal and ictal SPECT scans were analyzed using SISCOM, ISAS and STATISCOM to create hyperperfusion and hypoperfusion maps for each patient. The SPM based analysis was performed with a group of 30 control subjects. Perfusion maps were created with thresholds p = 0.001 for ISAS, p =0.027 for STATISCOM, and 2 standard deviations for SISCOM. A training set of 5 cases with strong temporal lobe activation were used to create consistent perfusion maps and to optimize thresholds. Mean ictal injection time (as determined by scalp EEG) was 26 s ± 8.7 seconds, and all were ictal injections. Hyperperfusion and hypoperfusion maps were co-registered to the patient's MRI. We created a custom software tool for displaying, reviewing, and annotating the thresholded hyperperfusion and hypoperfusion images in sagittal, axial a coronal planes. In this preliminary study, a single reviewer (VS) blinded to clinical data and analysis method (ISAS, STATISCOM or SISCOM) marked the lobe of probable seizure origin and rated his confidence (scale of 1 to 10) in the localization. Results: Out of 21 nonlesional temporal lobe cases ISAS correctly determined resection site in 19 (90%), STATISCOM in 18 (86%), and SISCOM in 13 (62%). Comparison of the three methods show a significant difference in localization between the SPM based methods to standard SISCOM, with p = 0.005 (ISAS vs SISCOM); p =0.011 (STATISCOM vs SISCOM). Mean confidence rating in ISAS cases was 7.26 (SD ± 0.96), in STATISCOM 7.30 (SD ± 1.43), SISCOM 4.43 (SD ± 2.5). Conclusions: Both SPM based methods of ictal and interictal SPECT processing were superior to the standard SISCOM analysis in terms of localization and confidence rating. There was no significant difference between ISAS and STATISCOM results. In the future we plan to: 1) Perform a multi-reviewer study to investigate the concordance between blinded expert reviewers for each analysis method. 2) Investigate the association between seizure outcome in temporal and extratemporal lobe epilepsy.
Neuroimaging