Information Mining in Brain Data

Brain functional connectivity, effective connectivity, and coordination among brain regions have become the hot problems in the studies of human brain functions and diseases. With more brain data accumulated, researchers in different fields are so eager to understand more profoundly how the brain systems work. For various brain data, scientists of information science have to face two basic problems: how to process exactly the brain data and how to mine hidden information in the data. In this paper, we introduce a few of multivariate statistical techniques used, such as principle component analysis (PCA), independent component analysis (ICA), structure equation model (SEM), dynamic causal model (DCM) and time-frequency analysis. But our emphasis would be mainly on the researches on some cognitive task conducted by our own group and give a few examples