Large-scale Functional Brain Network Research And Pattern Recognition Of Complex Partial Temporal Lobe Epilepsy

Complex partial temporal lobe epilepsy is a neurological disease characteriezd by partial discharge in the cerebral cortex,usually caused by ictal dischage from the temporal lobe system.The patients are usually accompanied with different degrees of disturbance of consciousness,and a minority of them may develop into a type of generalized seizures.Previous studies have shown that epilepsy is a disease in which the brain functional network is impaired,however,most of these studies focused on the intrinsic and extrinsic connections of the network and few studies focus on this specific type of epilepsy.Therefore,it is necessary to further explore the specificity of the topological properties of the functional network and the applicability of its application.Firstly,this paper introduces a large-scale functional small-world network model to explore the abnormalities of brain functional network properties caused by complex partial temporal lobe epilepsy.The study found that the brain functional network of patients with this disease had the small-world property as nomal subjects,but its network showed a decline in information transmission and processing capabilities.By analyzing the shortest path attributes of brain functional network of patients and normal subjects,we found that the damage of the patients' s network was maily due to the variability of the temporal lobe,parahippocampal gyrus,anterior central gyrus,basal ganglia,thalamus and amydala.These areas had significantly increased or decreased information transmission capability,resulting in decreased ability of the information transmission and processing capabilities of patients' s networks,which may lead to a series of clinical symptoms such as impaired consciousness and cognitive dysfunction.And we found that these altered network attributes were closely related to the patients' s clinical characteristics.Among them,the properties of the shortest path of left parahippocampal gyrus and the left basal ganglia were significantly negatively correlated with clinical characteristics,and the properties of the shortest path of left central anterior gyrus,right inferior temporal gyrus and the left thalamus all had a significant correlation with the age of onset.Further,considering the difficulty and the low efficiency of the diagnosis of complex partial temporal lobe epilepsy,we carried out a classification study by using the edges from functional networks,the network properties and the properties of the nodes from functional networks as features.The results showed that the classification model constructed by the edges and the properties of the node both had a good classification ability(82.1% of the classification accuracy rate),which means that they are effective features to identify the disease.The classification studies corresponding to the two types of features consistently revealed that complex partial temporal lobe seizures may affect the temporal lobe,motor area,frontal lobe and subcortical regions like basal ganglia,insula and cingulate gyrus.The two analysis methods in our studies make up for the insufficiency of the single analysis method.The studies based on changed topological properties of the functional network and the classification experiment together found that the damaged areas like cerebral temporal lobe,motor,parahippocampal gyrus,basal ganglia,insula and so on.And we found the potential imaging markers which may be used in diagnosing the disease.