独立分量分析及其在诱发电位提取中的应用

独立分量分析能够在各信号频率混迭的情况下,从它们的线性混合信号中有效分离出原始信号,这为微弱电生理信号的提取提供了途径,受到生物信号处理研究人员的关注.作者研究分析了独立分量分析模型和一种基于负熵判据的定点快速ICA算法的实现,并将该算法运用于仿真诱发电位的提取中,从观测信号(混合信号)中有效地提取出视觉诱发电位.

Independent component analysis and its application in estimating VEP

Original signals can be effectively estimated from their linear mixed signals by independent component analysis (ICA), even though their frequency components are overlapped, which provides a way to estimate the faint electrophysiological signals. And many researchers who study biomedical signal processing have paid great attention to the technique. We discuss the ICA model and the implement of a fixed-point fast ICA algorithm based on negentropy criterion. With the algorithm, we availably pick up the simulated visual evoked potential (VEP) from the mixing signals.