摘要
在脑电脑磁研究中 ,常将大脑视为一电磁系统。利用准静态近似下的麦克斯韦方程 ,可以发现代表脑神经元活动的原在电流密度与脑外磁场之间呈线性关系。在高斯噪声存在的情况下 ,采用极大似然估计理论 ,讨论了一种基于脑磁场时域 -空域数据求解磁源定位问题的一般方法。球对称导体模型下的模拟计算表明 ,这一方法是有效的。对于考虑真实头模型下的磁源定位问题求解 ,提出了一种联合使用脑磁脑电数据的近似方法。
The brain is often considered as an electromagnetic system on MEG/EEG studies. From quasistatic approximation of Maxwell's equations, it can be found that the relationship between the primary current density representing the cerebral neuron activities and magnetic field outside the head is linear. A general method, which can be used to locate the magnetic sources in the brain from spatio-temporal MEG data, is discussed. The method is based on maximum likelihood estimation, when the noise is Gaussian. Computer simulation using spherically symmetric conductor head model demonstrates that the method is effective. Considering the real head-shaped model, an approximate method based on the union of MEG and EEG data is also presented.
出处
《生物物理学报》
CAS
CSCD
北大核心
2000年第2期264-271,共8页
Acta Biophysica Sinica
关键词
脑磁图
磁源定位
极大似然估计
球对称导体模型
Magnetoencephalogram
Magnetic source localization
Maximum likelihood estimation
Singular value decomposition
