A Study Of Resting State-fMRI In Focal Epilepsy

Objective:To evaluate the feasibility of the resting-state fMRI based on ReHo, ALFF and fALFF for localizing epileptogenic zone in patients with focal epilepsy, and compare the diagnostic ability for different techniques (resting-state fMRI vs MRI,VEEG and FDG-PET) in epilepsy.Subjects and methods:Fifty-two patients with focal epilepsy who underwent comprehensive evaluation were consecutively included. A SIEMENS TRIO3.0T MR Scanner, was used to obtain conventional MRI sequences, high-resolution hippocampus anatomical images and resting-state fMRI data from all subjects. The data were preprocessed with Matlab7.1and SPM8. REST1.4software was used to generate the activation map of ReHo, ALFF and fALFF by dealing with the preprocessed data. The results were further analysed on the group and individual lever. The first part:RS-fMRI data were acquired from20unilateral mesial temporal lobe epilepsy patients (thirteen right-side, seven left-side, the seven right-side mTLE patients’images were flipped) and20age-and gender-matched healthy subjects, and were analyzed by using ReHo, ALFF and fALFF analysis at both whole brain and bilateral mesial temporal lobe lever. The second part:Fifty-two patients with focal epilepsy who underwent comprehensive evaluation were consecutively included. In order to compare the diagnostic value of MRI, VEEG, FDG-PET, and RS-MRI, the chi-square test were used to test their differences in diagnostic sensitivity, specificity, positive predictive value and negative predictive value. Those of ReHo, ALFF and fALFF were also compared.Results:1.①ReHo, ALFF and fALFF all can reveal brain regional functional abnormalities in left-side mTLE group;②the patient groups showed predominant increases abnormal brain activites in the mesial temporal lobe and the brain area ipsilateral to the epileptogenic side.2. The comparison results of MRI, VEEG, FDG-PET and RS-fMRI by the chi-square test:①the sensitivity difference between MRI and RS-fMRI were statistically significant, the sensitivity of RS-fMRI was higher than that of MRI(χ2 values were8.60, P values were0.003), but the sensitivity difference between VEEG, FDG-PET and RS-fMRI were no statistical significance (P>0.05).②the specificity, positive predictive value and negative predictive value differences between MRI, VEEG, FDG-PET and RS-fMRI were no statistical significance (P>0.05).3. The comparison results of ReHo, ALFF and fALFF by the chi-square test:①the sensitivity difference between ALFF, fALFF and ReHo were statistically significant, the sensitivity of ReHo was higher than that of ALFF and fALFF (χ2values were4.63,8.82, P values were0.0314,0.0030, respectively), but the sensitivity difference between ALFF and fALFF were no statistical significance (P>0.05).②the specificity, positive predictive value and negative predictive value differences among ReHo, ALFF and fALFF were no statistical significance (P>0.05).Conclusion:1. The resting-state fMRI based on ReHo, ALFF and fALFF shows a good feasibility. This technique may have the potential ability of detecting epileptogenic zone and network in patients with focal epilepsy.2. The resting-state fMRI based on ReHo, ALFF and fALFF has relatively high sensitivity and specificity for detecting epileptogenic zone. Resting-state fMRI shows a higher diagnostic ability than MRI and a comparable diagnostic ability to VEEG and FDG-PET which are uesd in current clinical realm, so it can be used in the research of clinical cases.