An Uncertainty Visual Analytics Framework for fMRI Functional Connectivity

AbstractAnalysis and interpretation of functional magnetic resonance imaging (fMRI) has been used to characterise many neuronal diseases, such as schizophrenia, bipolar disorder and Alzheimer ’s disease. Functional connectivity networks (FCNs) are widely used because they greatly reduce the amount of data that needs to be interpreted and they provide a common network structure that can be directly compared. However, FCNs contain a range of data uncertainties stemming from inherent limi tations, e.g. during acquisition, as well as the loss of voxel-level data, and the use of thresholding in data abstraction. Additionally, human uncertainties arise during interpretation due to the complexity in understanding the data. While existing FCN visual analytics tools have begun to mitigate the human ambiguities, reducing the impact of data limitations is an open problem. In this paper, we propose a novel visual analytics framework with three linked, purpose-designed components to evoke deeper interpretation of the fMRI data: (i) an enhanced FCN abstraction; (ii) a temporal signal view er; and (iii) the anatomical context. Each component has been specifically designed with novel visual cues and interaction to expose the impact of uncertainties on the data. We augment this with two methods designed for comparing subjects, by using a small multiples and a marker approach. We demonst rate the enhancements enabled by our framework on three case studies of common research scenarios, using ...
Source: Neuroinformatics - Category: Neuroscience Source Type: research