Research On Seizures Analytical Method Of EEG Signals

The study of EEG signals in patients with epilepsy mainly includes two parts: seizure prediction and seizure detection.Traditional on-off line prediction methods ignore the specificity of different epileptic subjects(e.g.,age and seizures region in brain)and the diversity of epileptic seizure modalities(e.g.,seizures during stress and awake state),which lead to low prediction accuracy and poor flexibility.Excessive EEG channel during automatic epilepsy detection increases the redundant features in the epileptic detection process,and the seizure detection model needs to use the artificially labeled seizure signal to complete the training,which makes the model adaptive effect poor.The main research contents are as follows:1.The thesis introduces the research background of seizure prediction and seizure detection,and focuses on the research status and research hotspots of seizure prediction and detection.Furthermore,it outlines the necessity of seizure prediction and seizure detection.Finally,the main work of the current research is described.2.Sparse group penalty algorithm and incremental learning mechanism are proposed to improve seizure risk prediction.In particular,sparse group penalty algorithm is proposed based on data correlation to incorporate the dependence structure among the features into solving approaches.Then,a relative spectral feature extraction approach is applied to construct a pattern library incrementally.Further,the prediction model parameters are dynamically updated and adjusted based on the updated subject-specific pattern library and incremental learning mechanism.The experimental results show that the proposed epileptic seizure prediction mechanism can sparse the parameters of the model and reduce the retraining time of the parameters.At the same time,it has high prediction accuracy and robustness,And the seizure focus determined by the model parameters can provide the doctor with a medical condition suggestion.3.A seizure detection mechanism with density clustering is proposed.Firstly,the EEG signal of the patient under normal conditions is analyzed.And the redundant EEG channel filtering is completed by using the standard deviation feature and the cross-correlation information,then use the mean filtering method and the difference method to smooth and detrend the extracted variance timing features.Next,thesis using the anomaly detection algorithm to complete the coarse positioning of the seizure,and finally using the density clustering method to cluster the abnormal points,and then complete the location of the initialand end position of the seizure.The simulation results show that the proposed seizure detection strategy is scalable and reduces the computational resource consumption,and there is no need to manually mark the seizure signal of the patient.The thesis proposes two strategies for seizure prediction and seizure detection.The former is based on feature dependence to achieve online seizure prediction.The latter is based on channel correlation and seizure signal and normal EEG signal to achieve adaptive seizure detection.Through simulation analysis,the accuracy of the model can be improved by considering the feature relationship.Finally,the research content and innovation points of the thesis are summarized,and the future research work is expected.