Faculty of Physics - University of Bucharest

Abstract:

In recent years, advanced non-invasive medical imaging techniques such as positron emission tomography, dynamic computer tomography, and magnetic resonance imaging (MRI) have been introduced into biomedical practice. Beyond the plain imaging of morphological structure, the analysis and visualization of biomedical image data is a challenge with growing importance for both basic research and clinical application. In this respect, functional MRI (fMRI), as a non-invasive technique in localizing dynamic brain processes in intact living brain, is by far the most complex and informative approach in neuroscience imaging. The present contribution critically evaluates the hypothesis-driven inferential methods in contrast with data-driven model-free techniques in the context of the analysis, processing, and interpretation of brain imaging data. The specificity of applying various exploratory methods to fMRI time series is highlighted and, consequently, their benefits and limitations are comparatively discussed and typified by experimental investigations reported in literature and by our own research as well. The emphasis is put on the independent component analysis (ICA) considered as a most promising exploratory multivariate approach to neuroimaging data analysis. ICA is based on a minimum of statistical assumptions on the latent sources and allows discovering features in fMRI data reporting on the organization of the nervous system. In this respect, we adapted some previously reported experimental protocols in a visuo-motor task that was inferentially analyzed only and which, subject of ICA, additionally revealed activity predominantly localized to auditory regions, with time courses of activation consistent with the experimental periods of fixation. Our results inferred that exploratory approaches allow discovering more activity in fMRI data than predicted in advance, so that additional regressors can complete a linear model with potential benefits in model accuracy and physiological interpretation of separated components.