Comparison of two methods of analysis on hand motor functional MRI: statistical method and independent component analysis

摘要

Here we analyze two methods to analysis on the hand motor functional MRI of human. Statistical methods, especially the correlation method, are considered as the golden standard for assessing where the region is with some statistical confidence. But, dependence on the fixed reference function can be a source of error and inflexibility to estimate variable HRF (hemodynamic response function). So we implemented the ICA as a data-driven methods free from prior knowledge and constraints. We found that both sets of results are almost the same in the sense of prior anatomical knowledge about the activation site of motor function.