Study Of The Localization Method Of Epileptogenic Focus In Children Based On Multi-mode Brain Image

Epilepsy is a chronic disease in which the sudden abnormal discharge of brain neurons leads to transient brain dysfunction.The incidence rate of epilepsy in children is about 5%,and about 30% of children do not respond to any medication.These patients are expected to become children without epilepsy after undergoing surgical resection of epileptic foci.Therefore,accurate localization of epileptic focus in children can not only improve the success rate of surgery,but also reduce the additional damage caused by surgery,which has far-reaching practical significance.In this paper,we first study the traditional based on EEG(electroencephalogram)source image location method.In this method,the minimum norm estimation(MNE)algorithm is used for source imaging of EEG;the relationship between EEG and brain source signal is clarified by solving the forward problem;in order to invert brain endogenous signal,the inverse problem is solved based on Tikhonov regularization,and finally the source is imaged to locate the epileptic focus.The simulation results show that when the signal-to-noise ratio of EEG is between 1-20 d B,the mean value of positioning error is between 10 mm-19 mm,the standard deviation of positioning is between 6-8,the positioning accuracy is low,and it is greatly affected by noise.Therefore,in this paper,aiming at the problem of low positioning accuracy in the existing methods,combined with the advantages of multi-modal brain imaging,such as MRI(magnetic resonance imaging),EEG,CT(computed tomography),s EEG(Stereoelectroencephalography)and so on,we propose a localization method of epileptogenic foci in children based on multi-mode brain image.Firstly,In order to obtain accurate EEG and s EEG position information and more suitable for children’s head model and source model,this paper studies multimodal brain image pretreatment method.Because of the confusion of MRI and CT original images,and the inconsistency of voxel and volume size,MRI and CT need to be reoriented and resampled;in order to eliminate the interference of EEG and s EEG measurement,we need to remove the interference;and because the brain structure of children is different at different ages,we have created a personalized head model and source model.MRI and CT are fused by rigid body transformation method,and Then,MRI,CT,EEG and s EEG are fused by means of correction or manual calibration to obtain the location information of EEG and s EEG.Thirdly,in order to improve the positioning accuracy,based on the EEG source reconstruction localization method,this paper introduces the s EEG measurement signal,and creates a positive problem model based on the s EEG and EEG.In order to clarify the relationship between EEG,s EEG and source signal,based on the boundary element method,the leadfiled matrix of EEG and s EEG is solved respectively,and then the forward problem model combined with EEG and EEG is established,and then the source signal is retrieved and the epileptic focus is located by solving the inverse problem.Finally,the simulation results show that under the same parameters,the mean value of the positioning error of the proposed method is 2.4mm and the standard deviation is 2.8,while the mean value of the positioning error of the traditional EEG based and EEG based methods is 16.94 mm and 3.8mm respectively,and the standard deviation is 16.7 and 3.8 respectively.At the same time,the positioning area of the proposed method is more accurate from the source imaging results and it is comprehensive,which verifies the effectiveness of the proposed method.