“School of Cognitive”
Back to Papers HomeBack to Papers of School of Cognitive
| Paper IPM / Cognitive / 7859 |
|
||||||
| Abstract: | |||||||
|
In this paper, a method based on the canonical correlation analysis (CCA) is developed for analysis of multi-subject fMRI data. The CCA produces a linear combination of fMRI data (across subjects), and a linear combination of bases of signal subspace, so as they have maximum correlation with each other. Since the proposed method is a multivariate analysis, it simultaneously uses time series of analogous voxels of all subjects and optimal bases of signal subspace. This in turn increases the flexibility through detecting different shapes of hemodynamic response in different regions and subjects. Using the proposed method for analyzing the simulated data illustrates its higher sensitivity in detecting active voxels. Application of this method to experimental fMRI data detects more activated regions than the general linear model (GLM).
Download TeX format |
|||||||
| back to top | |||||||
