基于单神经元放电记录技术和信号处理方法的癫痫发病预测研究

项目名称： 基于单神经元放电记录技术和信号处理方法的癫痫发病预测研究

项目编号： No.61201013

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 电子学与信息系统

项目作者： 李卓明

作者单位： 哈尔滨工业大学

项目金额： 25万元

中文摘要： 全球五千万的癫痫患者和癫痫突发的高危性使癫痫发病预测在过去30年中成为临床医生、神经生物学家、数学家、物理学家及工程师们一直合力挑战的研究课题。准确的癫痫发病预测使临床医生能及时地提供治疗或脑电刺激抑制癫痫发病。目前脑电信号的记录和分析被认为是预测癫痫发病的唯一手段，但是一直基于脑电图（EEG）记录的预测研究由于其信号反映的是大量同步神经元的电位和且分辨率较低，所以未达到可应用的准确预测率。以提高癫痫发病预测准确率为目标，本课题将利用一种新的、高分辨率、首次用于癫痫预测研究领域的多脑区、单细胞放电记录技术，通过数学、统计学和信号处理等方法比较癫痫发病前期的脑电信号和正常状态下的脑电信号差异，从而获得与癫痫发病相关的神经元异常放电信号并对癫痫发病进行准确的预测。本课题将基于这一能精确解码神经放电的记录技术建立一个准确可靠的癫痫发病预测模型，为癫痫发病预测的临床应用和电刺激治疗提供理论基础。

中文关键词： 癫痫癫痫发病预测；单神经元；局部电位；信号预处理；模式识别

英文摘要： Epilepsy is one of the most prevalent neurological disorders and seizure is one of the most dangers, affecting approximately 50 million people worldwide. Joint efforts from clinicians, neuroscientists, mathematicians, physicist, and engineers devoted to this very challenging epileptic seizure predication research field have been underway for the last 3 decades. Accurate seizure predication would enable clinicians to provide well-time treatments or electrical brain stimulation to inhibit the seizure onset. Currently, brain activity recording technique and data analysis are the only solutions to predict seizures. Up to now, seizure prediction has been based on EEG recordings which reflect postsynaptic potentials of a large number of synchronized neurons, but due to the low resolution of EEG data the accurate prediction of seizures still remains a challenging task to apply in real. In order to improve the accuracy rate of seizure prediction, our research will use a novel recording technique simultaneously collect single neuron spikes in multiple brain regions to combine with mathematical analysis, statistical methods and signal processing theory, which will compare neuron activities between pre-ictal and inter-ictal periods to search the specific signals to accurately predict the onset of seizures. This proposed st

英文关键词： seizure prediction；single unit；local field potential；signal pre-processing；pattern recognation