The diagnostic value of 18F-FDG PET/MRI in MR-positive refractory epilepsy

2021 
1598 Objectives: Refractory epilepsy accounts for at least one third of people with epilepsy. Surgery is an effective way to treat refractory epilepsy. The accurate localization of the epileptogenic zone (EZ) is very important before surgery. We mainly discuss the diagnostic utility of preoperative 18F-fludeoxyglucose (FDG) PET/MRI co-registration in refractory epilepsy which lesions are only found in the temporal lobe on the MR images. Methods: We performed a retrospective study that included 665 patients who underwent focal resection for drug-resistant epilepsy and follow up more than 12 months. There were 612 MR positive cases and 53 MR negative cases. In MR positive cases there were 306 cases localized in only temporal lobe on MR images. In these patients, they first did a MR scan, then a PET examination and 18F-FDG PET/MRI co-registration. Moreover, they also did a long-term scalp video-electroencephalogram (VEEG) monitoring. We located the EZ based on the results of MR, PET and VEEG. There were 215 patients underwent surgery directly, and 91 patients performed further stereoelectroencephalography (SEEG) examination before surgery. Results: In 306 patients, after combining PET/MR with VEEG, the accuracy of localizing EZ is 70.92%.There were 114 patients whose localization of MR, PET and VEEG were consistent with each other respectively, the accuracy of localizing EZ is 81.58%. 91 patients received SEEG examination, the accuracy of localizing EZ is 79.12%. Conclusions: 18F-FDG PET/MRI co-registration plays an important role in focal temporal lobe MR positive refractory epilepsy, PET/MRI co-registration combines with VEEG can accurately localize EZ. When MR,PET and VEEG are consistent with each other respectively, the patient can receive surgery directly without further SEEG examination.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []