Quantitative integration of dual-tracer PET and ictal SEEG refines candidate epileptogenic regions

To evaluate whether quantitative subregional analysis of dual-tracer PET (¹⁸F-FDG and ¹⁸F-FMZ), integrated with ictal stereoelectroencephalography (SEEG), improves localization of the epileptogenic zone and prediction of postoperative outcomes in drug-resistant epilepsy.

We retrospectively included 43 patients with drug-resistant epilepsy who underwent resective surgery. Subject-specific quantitative analysis was used to identify PET abnormalities, which were divided into three mutually exclusive subregions: FDG-only, FMZ-only, and their intersection (AND region). Within the actual SEEG sampling coverage, we assessed the imaging characteristics of each subregion, their spatial correspondence with the ictal epileptogenicity map (EM), and the associations between resection coverage and postoperative Engel outcomes.

The abnormal burden of the AND region was significantly higher in patients with Engel class I outcomes (P = 0.018). Although all PET subregions showed some spatial convergence with EM, only AND-EM concordance (P = 0.017) and its resection coverage (P = 0.016) were associated with favorable outcomes. EM intensity in the AND region was significantly higher than that in the FMZ-only region (P = 0.0039). The AND-EM fusion model achieved a cross-validated AUC of 0.902 for discriminating postoperative seizure freedom, with a cross-validated specificity of 0.874 at 80% sensitivity. In MRI-negative patients, it improved AUC by 0.22 over unimodal models.

Quantitative dual-tracer PET subregional analysis combined with ictal SEEG may help refine candidate epileptogenic regions and improve prediction of postoperative seizure freedom.