A framework for the automation of multimodalbrain connectivity analyses

Abstract

In neuroscience research, there has been an increasing interest in multimodal analysis, combining the strengths of unimodal analysis while reducing some of its drawbacks. However, this increases complexity in data processing and analysis, requiring a big amount of technical knowledge in image manipulation and a lot of iterative processes requiring user intervention. In this work we present a framework that incorporates some of this technical knowledge and enables the automation of most of the processing in the context of combined resting-state functional Magnetic Resonance Imaging (rs-fMRI) and Diffusion Tensor Imaging (DTI) data processing and analysis. The proposed framework presents an object-oriented architecture and its structure reflects the nature of three levels of data processing (i.e. acquisition level, subject level and study level). This framework opens the door to more intelligent and scalable systems for neuroimaging data processing and analysis that ultimately will lead to the dissemination of such advanced techniques.