Wireless EEG/PSG System with Novel Artifact Removal

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): EEG is a valuable non-invasive clinical tool in numerous applications, from the diagnosis and treatment of brain diseases to the clinical monitoring of neurological injuries, sleep disorders and depth of anesthesia. However, EEG signals are very susceptible to various artifacts which seriously impede the EEG interpretation and compromise its therapeutic capabilities. Methods currently employed for removing artifacts from EEG recordings are not clinically effective or feasible for real-time and long-term neuro-monitoring. Hence, the overall goal of this project is to develop a novel, high-fidelity artifact identification and removal technique that will be specifically useful for ambulatory EEG recording and intervention. The proposed novel artifact removal technique is based on the Wavelet-Based Artifact Removal (WBAR) method, which exploits the excellent time-frequency localization of artifacts provided by the wavelet decomposition. The WBAR method is computationally very efficient and allows for simultaneous, real-time removal of a variety of EEG artifacts. It has been recently developed by the PI and tested for a single EEG channel in an extensive clinical study as part of a novel depth-of-anesthesia monitor. The WBAR method will be improved by combining it with the Wavelet Neural Networks for the precise artifact classification, and recursive EEG Parameterization methods for the reliable estimation of the corrupted EEG components. The combination of these methods will result in fully automated, real-timeartifact removal technique that maximally preserves valid EEG information. The development and implementation of this novel method will greatly enhance the functionality and utilization of Cleveland Medical Devices' entire line of ambulatory wireless EEG/PSG systems.