The aim of this research initiative is to improve the diagnostic value of functional MRI data using a combination of high fMRI data quality and the power of data driven analysis methods and statistical criteria. The first aspect is achieved by a high magnetic field strength and ultra-fast data acquisition methods, the latter uses combination algorithms able of clustering, processing the data and validating the results. To obtain reliable/provable results it is necessary to develop and explore several analysis methods using simulated, hybrid and in vivo functional MRI data, i.e. using high field (3 Tesla) single-shot ultra-fast EPI. The theoretical and experimental analysis will be carried out by means of both standard and combination strategies. The standard strategies will feature both model-based (t-tests, correlation analysis) and explorative model-free, data driven methods as principal component analysis (PCA) and independent component analysis (ICA),factor analysis (FA), or unsupervised pattern recognition, neural network based approaches (crisp and fuzzy clustering analysis). These systematic validation steps will thus lead to more reliable results in the evaluation of fMRI time series.