论文部分内容阅读
独立成分分析（ICA）是信号处理领域中新近发展起来的一种很有应用前景的方法,而脑功能磁共振（fMRI）信号的有效分离与识别是一个正在研究和实验之中的技术领域,因此,发展基于ICA的fMRI数据处理方法具有明显的理论价值和应用前景.首先分析了现行ICA-fMRI方法采用的信号与噪声的空域分布相互独立的信号模型所存在的明显不足,然后提出了微域中的信号与噪声的时域过程相互独立的fMRI信号模型,从而建立了一种新的fMRI数据处理方法:邻域独立成分相关法.从理论和仿真实验两个方面阐明了新方法的合理性,最后给出了实际fMRI数据的例子.
Independent Component Analysis (ICA) is a newly developed method in the field of signal processing. However, effective separation and identification of fMRI signals is a field of research and experiment. Therefore, the development of ICA-based fMRI data processing has obvious theoretical value and application prospects.Firstly, the existing shortcomings of the current signal-noise independent spatial signal distribution model used in ICA-fMRI method are analyzed. Then, Domain signal and noise time-domain process independent fMRI signal model, thus establishing a new fMRI data processing method: neighborhood independent component correlation method.From both theoretical and simulation experiments to clarify the reasonableness of the new method Finally, an example of actual fMRI data is given.