区域增长作为一种有效的功能磁共振(fMRI)分析方法,由于受到诸如噪声、生长准则等因素的影响,限制了它在脑功能激活区检测方面的应用与发展.为了克服这些问题,提出一种自适应的区域增长方法,用于fMRI脑功能激活区的检测.该方法首先利用主成分分析(PCA)对预处理后的图像进行降噪;然后通过裂分合并与模板匹配相结合的方法自动获取初始生长点;最后使用典型相关系数和皮尔森相关系数作为生长准则进行区域增长.通过模拟数据实验和真实数据实验验证了该方法的有效性.实验结果表明,相比于其他的fMRI分析方法(如ICA、SPM和基于裂分合并的区域增长方法),自适应区域增长方法能够获得更加准确有效的结果.而且该方法还能应用到静息态数据的分析中,进一步证明了该方法的可行性与有效性.综上,自适应区域增长方法能够拓宽区域增长在功能磁共振数据分析中的应用.%Region growing has been utilized in the analysis of functional magnetic resonance imaging (fMRI) data for many years, while some influential factors, such as the noise problem, definition of the homogeneity criterion, restricting the application and development in brain function active region detection.In order to overcome these disadvantages, an adaptive region growing method (ARGM) is proposed to detect fMRI brain function active region, where PCA was firstly used to de-noise the fMRI data as a step of preprocessing.Then the region seed was automatically selected by the split-merge method combined with a prior template.Next, an improved homogeneity criterion defined by Canonical correlation coefficient and Pearson correlation coefficient were used to judge the region growing.Compared with the typical fMRI data analysis methods such as ICA and SPM and the classical region growing method, ARGM generates a more accurate and reliable result in task-related experiment.In addition, the resting-state experiment has also demonstrated the effectiveness and usefulness of the proposed method.To conclude, the proposed method is able to broaden the application of region growing in analyzing fMRI data.
展开▼
