| Temporal lobe epilepsy(TLE)and generalized epilepsy(GE)are two common types of epilepsy,and their etiology and pathogenesis are different.Temporal lobe epilepsy refers to seizures originating in the temporal lobe region,and its seizure mechanism is related to abnormal neural electrical activity in the hippocampus and parahippocampal gyrus,and may also be related to factors such as abnormal synaptic transmission of neurons and neuron damage.Generalized epilepsy refers to a kind of epilepsy characterized by generalized seizures without obvious reasons.This article discusses the difference between temporal lobe epilepsy and generalized epilepsy from the perspective of functional brain network,using scalp EEG data after clinicians judge and classify the types of epilepsy patients,and combining network analysis methods to analyze the network neural mechanism of epilepsy patients.The research content mainly includes energy analysis,functional brain network topological connection analysis and network attribute analysis,and using the above differences to identify epilepsy classification,so as to reveal the potential application value of functional brain network in the clinical diagnosis and treatment of epilepsy.The specific research content of this paper is as follows:1.Head-surface network research in temporal lobe epilepsy and generalized epilepsy.In cooperation with Sichuan Provincial People’s Hospital,the 24-hour(Electroencephalogram,EEG)signals of 50 patients with epilepsy were collected,including 20 cases of TLE patients and 30 cases of GE patients.We selected the 10-minute EEG data of each patient in the awake state.In this chapter,we first explore the brain differences between temporal lobe and generalized epilepsy in terms of energy and brain complexity and flexibility.Specifically,we use power spectral density(PSD)to calculate brain energy,while fuzzy entropy(FE)is used to measure brain complexity and flexibility.The results showed that in the β(13-30Hz)and γ(30-45Hz)frequency bands,the PSD of generalized epilepsy was significantly larger than that of temporal lobe epilepsy in almost the entire brain area,while the FE of generalized epilepsy was mainly in the frontal lobe,central and regional regions.The FE of temporal lobe epilepsy is mainly strong in the temporal lobe.Then,we constructed and analyzed the corresponding network topologies respectively.The results showed that the network topology of generalized epilepsy showed stronger fronto-parietal-occipital lobe long-range connections in different frequency bands,while temporal lobe epilepsy mainly showed stronger fronto-centrotemporal lobe short-range connections in the low frequency bands,namely the δ and θbands.Based on the constructed coherent network,we calculated four types of network properties,clustering coefficient,characteristic path length,global efficiency and local efficiency,to further quantify the network differences between the two types of epilepsy.Consistent with the network topology results,generalized epilepsy exhibits stronger network properties in the high-frequency bands,that is,the β frequency bands,which are manifested in larger clustering coefficients,global efficiency and local efficiency,and smaller characteristic path lengths.Finally,we use PSD value fusion network features as classification indicators to identify epilepsy classification.The results showed that the classification accuracy was 82%.The abnormal brain areas of temporal lobe epilepsy and generalized epilepsy are different,in the left and right brain and the front and rear brain areas respectively,this study may provide potentially effective biological indicators for epilepsy research and clinical auxiliary diagnosis and treatment,and help reveal the underlying neural mechanisms,especially network mechanism differences,of different types of epilepsy.2.Source space network studies of temporal lobe epilepsy and generalized epilepsy.In this section,we perform source localization on EEG data for temporal lobe epilepsy and generalized epilepsy,respectively,to obtain cortical time series.Subsequently,we first calculated the PSD of the two types of epilepsy and made statistical analysis,and then further constructed the source space network and made statistical analysis.In the present study,we performed power spectral analysis on 6239 voxels after tracing the origin of temporal lobe and generalized epilepsy.It was found that the strong activation areas of the temporal lobe mainly appeared in the δ frequency band,and concentrated in the temporal lobe,parietal lobe,and occipital lobe;while the strong activation areas of generalized epilepsy were distributed in 5 frequency bands,mainly in the limbic lobe,frontal lobe,and parietal lobe.There is also obvious activation in the temporal lobe area of the γ frequency band.Then the coherence network was also constructed,and the results showed that most of the strong connections in temporal lobe epilepsy were located in the hippocampus and adjacent medial temporal lobe structures,and the strong connections in generalized epilepsy almost spread throughout the entire brain region,especially in the frontal lobe,occipital lobe and thalamus,and it is closely connected with the part of thalamus related to cognitive function.Source space results suggest temporal lobe epilepsy may be associated with abnormalities in hippocampal regions,and generalized epilepsy may be associated with abnormalities in the frontal lobe and thalamus.To sum up,this paper found the differences in the network mechanism of temporal lobe epilepsy and generalized epilepsy through the analysis of head table and source space EEG network respectively,and revealed the unique advantages of EEG network in mining the disordered brain information interaction mode of epilepsy patients.which contributes to the development of clinical auxiliary diagnosis methods and treatment methods for epilepsy. |
PDF Full Text Request