Comparative Study Of Multi-model Functional Image Of MR And PET-CT In The Localization And Orientation Of Temporal Lobe Epileptic Focus

ObjectiveTo evaluate the efficaciency of magnetic resonance imaging(MRI),magnetic resonance spectroscopy(MRS), rest-functional MRI and18F-FDGPET-CT brain imaging in the localization and Orientation of temporal lobeepileptic focus. And to evaluate their sensitivity, specificity, and accuracy in thediagnosis of temporal lobe epilepsy location through contrast withpostoperative pathology.Methods17cases of TLE patients all underwent conventional MRI, MRS, rest-fMRIand18F-FDG PET-CT brain imaging. Firstly, bilateral hippocampal volume andthe ratio of NAA/(Cho+Cr) were obtained by Siemens workstation. Andbilateral hippocampal signal intensity was observed in the oblique coronalsections to determine the epileptogenic side. Then the rest fMRI images werepostprocessed by SPM5software. Whole brain ReHo values were calculatedas well as the ReHo values of the both side of hippocampus. PET-CT imageswere analyzed by visual analysis method and Asymmetry index method todetermine the side of epileptic foci. Two independent samples t-test, Receiveroperating characteristic (ROC) curve and Pearson correlation analysis wereused for data analysis. Finally, the multi-model diagnostic results werecompared with the pathological results, and the efficiency of MRI, MRS and18F-FDG PET-CT brain imaging were evaluated in localization of temporal lobeepileptic focus.Results1. The ipsilateral hippocampal volume in TLE patients was significantlysmaller than the contralateral and normal control group(p<0.05). In controlgroup, the right hippocampus volumes were slightly larger than the left hippocampal volume, but their was no significant difference(p>0.05). Inaddition, there was no significant correlation between the ipsilateralhippocampal volume and the course of disease.2. The ratio of NAA/(Cho+Cr) and ReHo values in ipsilateral hippocampalin the TLE group were less than contralateral hippocampal and control group (ttest, p <0.05). When the NAA/(Cho+Cr) was0.415, the sensitivity andspecificity of the MRS to the diagnosis of HS were70.6%and93.8%. WhenReHo value was0.36, the sensitivity and specificity of the diagnosis of HSwere88.2%and94.1%.3. The diagnosis positive rate in TLE by PET-CT was100%and all patientsshowed low metabolism in the temporal lobe.4. There was good consistency between the diagnosis of PET andrest-fMRI and the kappa value was0.49.The ipsilateral hippocampal SUV waslower than the contralateral hippocampal SUV.5. Compared to pathology of removal tissue after surgery，the accuracy forlocalition of epileptic focus by conventional MRI, MRS, rest-fMRI and18F-FDGPET-CT brain imaging were64.7%,76.47%,70.6%and64.7%, respectively.When the four techniques were combined,we could further improve theaccuracy for localization of temporal lobe epileptic focus.Conclusion1. Among the four techniques, PET-CT gets the highest positive rate andMRS gets the highest diagnostic accuracy.2. Combined use of multi-model techniques can significantly increase theaccuracy and reliability of the localization of temporal lobe epileptic focus.