Presurgical Comprehensive Evaluation And Investigation Of High-frequency Oscillations In Source Localization And Brain Network Based On Magnetoencephalography In Patients With Insular Epilepsy

Insular epilepsy is complicated because insular seizures may mimic frontal,temporal,and parietal lobe seizures.The clinical complexity and variation in insular seizures frequently results in a failure to detect the foci in the insular cortex responsible for some surgical failures.Notably,the insula itself and its extensive connections with other potential epileptogenic zones(EZ)can result in variability of semiology and diagnostic uncertainty or mistakes.Therefore,it is a challenge to diagnose insular epilepsy based on clinical symptoms alone.Intracranial electroencephalography(ICEEG),especially stereo-electroencephalography(SEEG),has been demonstrated to define insular epileptic discharges and EZ through anatomo-electro-clinical correlations.ICEEG recordings are extremely invasive and risky;even so,only a few contacts end up sampling the insula itself,and extensive coverage is rarely possible.Magnetoencephalography(MEG),a relatively new noninvasive investigation technology with better spatial resolution than electroencephalography(EEG)and better temporal resolution than functional magnetic resonance imaging(fMRI),has shown the ability to localize epileptogenic foci.Unfortunately,only a few studies have reported MEG single equivalent current dipole(sECD)applied to the study of insular epilepsy.The discovery of high-frequency oscillations(HFOs,>80 Hz)in the brain opens a new avenue for the study of intrinsic brain abnormality in epilepsy because HFOs are more localized than low-frequency spike-and-wave discharge and could serve as a biomarker of seizure-onset zones.Nevertheless,previous reports on epilepsy HFOs have mainly used ICEEG recordings.The characteristics of HFOs on MEG in epilepsy,especially those of high-frequency neuromagnetic signals in insular epilepsy,remain largely unexplored.Recent studies that have investigated structural and functional networks have revealed that human brain is a complex network.In terms of epilepsy,the idea that “focal” epilepsies are not in fact so focal,but involve networks of varying scales,consequently,have led to a shift from “focus” to “networks”.Epileptogenic networks are defined by the brain regions involved in the production and propagation of epileptic activities.Network analysis has been applied to improve our understanding of chronic epilepsy,as it has provided valuable information on seizure onset,propagation,and termination,on the interictal state of functional networks in epilepsy and on alterations in structural networks.Based on the above research background,in the current study,we summarized the semiology and evaluated the value of noninvasive sECD of MEG data in the localization of EZ.Subsequently,high-frequency neuromagnetic signals on MEG were analyzed to evaluate their value in the source localization of EZ and epileptic networks in patients with insular epilepsy.Part One Insular Epilepsy: Semiology and Presurgical Localization InvestigationsObjective: To summarize the semiology and to evaluate the preoperative localization methods in patients with insular epilepsy.Methods: The data of patients with insular epilepsy who were identified by resective surgery from January 2010 to December 2016 were retrospectively analyzed.The role of semiology,scalp EEG,MRI,MEG and ICEEG in the localization of epileptogenic zones for patients with insular epilepsy was evaluated.Results: Twenty-three patients met the inclusion criteria.There were nineteen patients with auras,including six patients with somatosensory auras,three with viscerosensory auras,seven with vegetative or psychic auras,one with both somatosensory and viscerosensory auras,one with both somatosensory and vegetative auras,and one with both gustatory and viscerosensory auras.Of fifteen patients with motor symptoms,seven developed hypermotor seizures(HMS),including four with type I HMS,two with type II,and one showing features of both type I and type II.Six patients had somatomotor and visceromotor symptoms,one suffered visceromotor seizures and one experienced aphasia.Scalp EEG had no localization value in insular epilepsy.Interictal scalp EEG abnormalities showed lateralization in 13 patients and ictal EEG in 15 patients.Four of 23 patients with nontumoral insular epilepsy had insular abnormality on MRI.MEG spikes were visually identified in 21 patients.Dipole clusters of those spikes were localized in the insula or insular operculum in 15 patients(71%),and MEG showed better concordance than MRI in the localization of insular epilepsy.Interictal ICEEG showed insular or opercular epileptic activity in 14 patients(67%),and ICEEG captured seizures arising from the insula or insular operculum in 16 patients(76%).There was no statistically significant difference in insular epileptogenic zone localization between MEG and ICEEG.Histopathological analyses of the resected epileptogenic insula in 20 of 23 patients revealed focal cortical dysplasia.Seventeen patients(74%)were seizure free(Engel class I)after surgery.Summary: Noninvasive MEG may provide key supporting information for the presurgical localization of insular epileptogenic zone.The combination of semiological features and findings from noninvasive localization tests can help identify,diagnose and thereby better manage insular epilepsy.In nontumoral epilepsy surgery,the most frequently identified lesions are focal cortical dysplasia.A tailored resection can lead to an excellent seizure outcome.Part Two Detection and Localization of Interictal Ripples with Mag-netoencephalography in the Presurgical Evaluation of Drug-resistant Insular EpilepsyObjective: Precise noninvasive presurgical localization of insular epilepsy is of great significance.The objective of the present study was to detect and localize interictal HFOs in patients with insular epilepsy at the source levels using MEG and to investigate whether HFOs could delineate EZ by correlating the resection of HFOs-generating areas with surgical outcome.Methods: MEG data were analyzed with new accumulated source imaging(HFOs,80-250 Hz ripples during spikes)and conventional dipole modelling(spikes)methods for localizing epileptic foci.The relationship of the resection of focal brain regions containing interictal HFOs and the spikes with the postsurgical seizure outcome was analyzed.Results: Interictal HFOs were localized in the insular epileptogenic zone in 18 out of 21 patients undergoing surgical treatment for clinically diagnosed insular epilepsy.However,dipole clusters of spikes were involved in the insular EZ in 15 patients.Both the HFOs and the dipole cluster were localized in the insula in 14 patients.The seizure-free percentage was 87% after the resection of brain regions generating HFOs,whereas 80% after the resection of brain regions generating spikes.There was a much higher chance of freedom from seizures with complete resection of the HFOs-generating regions than with partial resection or no resection(P=0.031).No such difference was found for spike-generating regions.Summary: To our knowledge,this study is the first one to disclose the characteristics of HFOs on MEG in insular epilepsy.Our results suggest that HFOs from insular epilepsy could be noninvasively detected and quantitatively assessed with MEG technology.MEG HFOs(ripples during spikes)may be of great value for the localization of the epileptogenic zone in insular epilepsy.Part Three Altered Effective Connectivity Network in Patients with Insular Epilepsy: A High-Frequency Oscillations Mag-netoencephalography studyObjective: The objective of the present study was to investigate whether the effective connectivity(EC)network of patients with insular epilepsy is altered during the interictal period compared with healthy controls based on HFOs(ripples,80-250 Hz)in MEG data.Methods: The MEG data from 22 patients with insular epilepsy and 20 healthy controls were analyzed.Correlation and Granger causality analyses were performed to construct an EC network at the source level using resting-state HFOs from the healthy controls and interictal HFOs with and without spikes from patients with insular epilepsy.Alterations in the spatial pattern and connection properties of the patients with insular epilepsy were investigated in the entire brain network and insula-based network by comparing the patients with the controls.Results: The EC predominantly involving the insular cortex was significantly increased in the patients with insular epilepsy during interictal HFOs with spikes.Analyses of the parameters of graph theory revealed the over-connectivity and small-world configuration in the global connectivity patterns observed in the patients.In the insula-based network,the insular cortex ipsilateral to the seizure onset displayed increased efferent and afferent effective connectivity.Left insular epilepsy displayed strong connectivity with the bilateral hemispheres,whereas right insular epilepsy showed increased connectivity with only the ipsilateral hemisphere.Summary: This study provides neuromagnetic MEG data showing that patients with insular epilepsy display alterations in the EC network that differ from healthy subjects in terms of both whole-brain connectivity and the insula-based network during interictal HFOs.Alterations in the HFOs-based EC network provide distinct insights into the complex pathophysiological mechanisms of insular epilepsy,which may prove to be a novel promising biomarker of this type of epilepsy.Conclusions:1.MEG may provide key supporting information for the presurgical localization of insular epileptogenic zone.The combination of semiological features and findings from MEG can help identify,diagnose,and thereby better manage insular epilepsy.2.Interictal HFOs from insular epilepsy could be noninvasively detected and quantitatively assessed with MEG technology.MEG HFOs may be valuable for the localization of the epileptogenic zone in insular epilepsy.3.The patients with insular epilepsy display interictal alterations in the EC network that differ from healthy subjects in terms of both whole-brain connectivity and the insula-based network based on HFOs MEG data.Alterations in the HFOs-based EC network provide distinct insights into the complex pathophysiological mechanisms of insular epilepsy,which may prove to be a novel promising biomarker of this type of epilepsy.