The Design Of Mobile Monitoring System For EEG In Epilepsy

Epilepsy characterized by episodes of recurrent and unpredictable is a serious disorder of the central nervous system. Electroencephalogram(EEG) signals is an important way of monitoring and diagnosis of epilepsy. However, the traditional EEG monitoring can only be done in hospital, is not conducive to daily monitoring in patients with epilepsy, and due to the complexity of EEG signals, pose great challenges for artificial diagnosis of epilepsy.This paper proposes a comprehensive set of the mobile medical monitoring system, portable acquisition and transmission of EEG signal and according to the information for automatic analysis and detection of seizures. The system includes two main modules: front-end module and back-end module.The front-end module is a portable device based on ADS1299, including four aspects about front-end module in this paper: signal acquisition, control, data storage and data transmission. This article abandoned the previous EEG acquisition equipment necessary for the design of analog and digital circuit, instead of using the relatively advanced ADS1299 chip, using it to EEG amplification, filtering, sampling and other various operations. This set of acquisition equipment without the use of high frequency and high performance control chip, instead of using the data processing program into the computer, so it only need to use the lower frequency of MCU, it not only maintains the proper function, but also reduce the price of the whole system, and give full play to the advantage of the back-end module. In connection with the back-end module, taking into account the characteristics of the portable device, the device abandons the traditional USB wired connection, instead of using WiFi, it can easily connect with many types of freedom of the host computer, extended effective connection distance. In power, the device uses a very popular and common in mobile power, which can effectively eliminate the 50 Hz frequency interference, and further enhance the stability of the equipment.There back-end module includes preprocessing, feature extraction, feature normalization, feature selection and classification. Based on related epileptic EEG data in a public database as an example, first of all, the raw EEG signals is processed by bandpass and notch filtering; and then wavelet decomposition, and then feature extraction based on wavelet energy entropy, finally, the support vector machine(SVM) signals can be divided into non ictal and interictal state and complete seizure detection. The simulation results prove the effectiveness of the proposed scheme.