てんかんのMRI陰性例における局在診断の改善：デュアルプローブPET/MR（18F-FDG/11C-FMZ）の相乗効果

The localization of refractory epilepsy, especially MRI-negative cases, is a critical challenge in diagnosing and treating neurological disorders. To address this challenge, this study proposes a standardized protocol for localizing the epileptogenic focus using a dual-probe Positron Emission Tomography/Magnetic Resonance Imaging (PET/MR) technique. The protocol focuses on radiotracer selection, imaging sequence design, and image interpretation strategies. Radiotracers selection: This imaging combines ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) metabolic imaging and ¹¹C-flumazenil (¹¹C-FMZ) GABA_A receptor imaging to locate epileptogenic focus in refractory epilepsy patients by detecting glucose metabolism abnormalities and GABA_A receptor density changes; imaging sequence design: FDG PET/MR scans were done first, followed by FMZ PET/MR scans 24 h later, MRI Sequence images including structural and functional images; image interpretation strategy: The dual-probe evaluation strategy identifies potential epileptogenic focus through FDG hypometabolism and FMZ binding reduction, with concurrent findings providing strong evidence, quantitative analysis involves standardized uptake values (SUV) and asymmetry index (AI) of FDG and FMZ images. This study details the preparation process of radiotracers, scanning parameters, and image fusion methods, validating the protocol's effectiveness through representative results. The adoption of dual-probe PET/MR imaging technology may enhance the accuracy of epileptogenic focus localization, thereby supporting more precise assessments and improving treatment outcomes. Currently, this protocol has completed methodological validation, and preliminary results indicate its potential to guide preoperative assessment for clinical epileptogenic focus resection (successful surgery was performed in 3 patients based on localization results, with no post-operative seizures). To address technical limitations (e.g., short ¹¹C-FMZ half-life, cyclotron dependency), optimization directions include developing long-half-life analogues and exploring multi-tracer combinations. Future integration with AI-assisted image analysis and lesion identification is also feasible. In summary, this study offers novel insights for precision epilepsy diagnosis and treatment, holding significant implications for improving patient outcomes.