Recent advances in the data analysis method of functional magnetic resonance imaging and its applications in neuroimaging

摘要

Functional magnetic resonance imaging (fMRI) has opened a new area to explore the human brain. The fMRI can reveal the deep insights of spatial and temporal changes underlying a broad range of brain function, such as motor, vision, memory and emotion,all of which are helpful in the clinical investigation. In this paper, we introduce some recent-developed algorithms for fMRI signal detection such as model-driven method (general linear model, deconvolution model, non-linear model, etc.) and data-driven method (principle component analysts, independent component analysis, self-organization mapping, clustered constrained non-negative matrix factorization, etc.). We also propose several important applications of neuroimaging and point out their shortcomings and future perspectives.