摘要
目的:癫痫是以脑内神经元异常放电致部分或整体脑功能障碍为特征的慢性疾患。应用独立分量分析技术对癫痫高危人群进行神经系统电生理筛查,及早发现和对相关人群进行干预,减少癫痫的受累程度和致残、致死率。方法:建立独立分量分析的数学模型,使用自行研制的脑电信号采集仪采样16导标准脑电信号,并通过快速独立分量分析(FastICA)算法对数据进行处理。结果:处理16导标准脑电信号,分离出癫痫特征波,并对特征波进行识别,从而得到对癫痫的诊断;在此基础上将癫痫特征波反映射到16导标准电极,应用相关源电位软件对癫痫灶进行初步定位。结论:应用独立分量分析技术来实现癫痫的规模筛查,不仅降低了癫痫信号识别的工作量,提高了识别效率和正确率,使癫痫的筛查成为可能,而且可以根据识别结果对癫痫放电信号进行相关逆变换,再利用相应的源定位软件确定癫痫灶的位置,提高源定位的精度和效率。
ysis （ICA）, Objective ： Scanning the latent epileptic under the technology of independed component analfinding the patients and intervening in course of diseases earlier, reducing the degree of implicating and the ratio of disability or dead. Methods：Establishing ICA mathematic model, Sampling 16 channels standard EEG using the instrument developed by ourself, and processing the data using Fast ICA algorithm. Results：Separating, recognizing epileptic waveform, and diagnosing epilepsy by the resuits, and reversed transforming the epileptic waveform into 16 channels EEG space and locating the epileptic focus using some locating software. Conclusion ： Scanning epilepsy under ICA, not only reducing signals recognizing workload, but also increasing its efficiency, and makes it possibe to scan epilepsy; and the accuracy and efficiency of source location could be increased greatly by inverse ICA.
出处
《医学研究生学报》
CAS
2007年第3期278-280,共3页
Journal of Medical Postgraduates
基金
江苏六大人才高峰重点课题基金资助项目(批准号:2005A5)
南京军区南京总医院科研基金资助项目(批准号:2006068)
关键词
独立分量分析
脑电图
癫痫
源定位
Independent component analysis
Electroencephalogram
Epilepsy
Source location
