Based EEG And ECG Childhood Epilepsy Syndrome Analysis

Epilepsy syndrome is a precise type of epilepsy disease,which contains a variety of complex seizure symptoms.The precise classification and analysis of epilepsy syndrome enables doctors to take proactive preventive measures in time and use anti-epileptic drugs in a targeted manner.At present,the classification of childhood epilepsy syndrome only relies on expert experience and clinical features of electroencephalogram(EEG).But it is still faced with some problems,including: 1)Lack of research on the classification algorithm of children’s epilepsy syndrome with multimodal signals;2)Lack of accurate classification of different epilepsy syndromes between seizures and interictal periods;3)Lack of correlation research and statistical analysis of epilepsy syndrome,different characteristics,and different periods.In response to these problems,the main research work of this dissertation is divided into the following aspects:1.A classification algorithm for childhood epilepsy syndrome based on the fusion of electroencephalogram(EEG)Mel-Frequency Cepstral Coefficients(MFCC)and Linear Predictive Cepstral Coefficients(LPCC)features,and electrocardiogram(ECG)heart rate variability(HRV)features is proposed.This study mainly focuses on the classification models of two common childhood epilepsy syndromes,CAE and WEST.This method uses discrete wavelet transform technology to remove the baseline drift in the original ECG and perform Rwave positioning,and further calculate the time domain and frequency domain characteristics of EEG and HRV.In order to solve the problem of sample imbalance between the onset and between attacks,the SMOTE+Tomek Links data balance method was studied to synthesize minority samples.Through multiple sets of experimental contrast,the proposed EEG+ECG fusion feature classification method for childhood epilepsy syndrome has reached 98.37% accuracy on the CHZU data set.It is verified that the classification performance of multi-modal physiological signals is better than that of single physiological signals.2.A classification algorithm for childhood epilepsy syndrome based on statistical analysis and optimization of ECG HRV features and EEG features is proposed.This study is mainly aimed at the two periods of three epilepsy syndromes(CAE,WEST,and BECT)(seizure period and interictal period).To study epilepsy syndrome,ECG’s HRV and EEG characteristics,and the correlation between different periods(seizure period and interictal period).Compared with the traditional attention to the static characteristics of HRV signals,this method adds nonlinear dynamic characteristics to characterize the dynamic characteristics of cardiac electrical activity.According to the time domain,frequency domain,and nonlinear characteristics of HRV in different periods,a one-way analysis of variance was carried out,and the correlation between the HRV of ECG and EEG characteristics in different epilepsy syndromes was discussed.Finally,the analysis results are used as the theoretical basis for feature selection,and the performance of the classification model of childhood epilepsy syndrome based on EEG,ECG,EEG+ECG features is studied through multiple sets of comparative experiments.It proves the validity of the proposed statistical analysis and syndrome classification algorithm based on EEG and HRV features of ECG.3.An intelligent classification system based on EEG+ECG for childhood epilepsy syndrome was developed.An online real-time classification platform is built through MATLAB GUI software.The system includes four sub-modules: signal preprocessing,R-wave positioning,feature extraction,and classification and recognition.It provides convenience for assisting doctors in clinical classification and targeted treatment.