Study On Epilepsy EEG Signal Localization Algorithm Based On Cyclic Spectrum And Sample Entropy

Epilepsy is a clinical manifestation of paroxysmal abnormality in brain neurons, has the characteristics of repetitive, sudden and temporary. As an important tool to study the characteristics of epileptic seizure, EEG is not provided by other physiological methods. Signal processing and pattern recognition methods of epileptic EEG signals fast and accurate positioning, is of great significance to reduce the burden on doctors and improve the efficiency of diagnosis of epilepsy.According to the current research at home and abroad epileptic EEG localization algorithm has made some progress, there is also some problems:1) EEG signal belongs to complex micro volt non-stationary signal, extremely vulnerable to the influence of interference signals, the traditional feature extraction algorithm cannot guarantee the validity of; 2) the EEG signals are huge amount of data processing in general algorithm in real time is not high; 3) alone a signal feature implementation of epileptic EEG localization may will misjudgment, positioning accuracy needs to be improved.This paper presents a method based on cyclic spectrum and sample entropy analysis for the detection and localization of epileptic EEG signals, in order to meet the localization of epileptic foci in patients with high risk of epilepsy in clinical medicine. Using principal component analysis method for data compression, solve the huge amount of data processing problems of multi-channel EEG processing and entropy analysis method at the same time, to ensure the real-time performance of the algorithm; the frequency notch filter and FIR digital filter to collect EEG signal preprocessing to reduce the impact of harmonics and other interference on location and to ensure the effectiveness of the algorithm; using principal component analysis method for data compression, real-time algorithm; cyclic spectrum density is greater than the normal EEG signals of epileptic EEG signals, can make a preliminary localization of the epileptic EEG; entropy analysis as the classic nonlinear dynamic analysis method, can effective features of brain electric signal. Combination of two methods to achieve thefinal EEG signal fast and accurate positioning. The experimental results show that the algorithm has good accuracy and sensitivity, and it is helpful to the localization of clinical epilepsy. The preoperative localization of epileptic focus is important.