Data Driven Exploration of Network Connectivity in task-fMRI of Focal Hand Dystonia

Background: Dystonia is a network level brain disorder with limited understanding of interactions between functional networks. Prior studies have demonstrated abnormalities both between and within functional networks. However, these findings were based on a hypothesis driven region of interest (ROI) approach or used resting state fMRI to extrapolate network differences in task-specific dystonia called writers cramp (WC). As a result, it remains unclear if these same networks would be active during the actual performance of a dystonic task. It9s also unclear from prior studies which networks play a role in the primary disease versus a compensatory region. Lastly, neuroimaging studies have used sequencing tapping task with equal motor performance to study networks changes due to concern that differences in motor performance during the dystonic writing task may confound the findings. Objective: The aim of this research study was: 1) to use a data driven approach on a task-specific focal dystonia to understand network level interactions during task-fMRI, 2) to determine if differences in observed network level could distinguish between brain regions of primary vs compensatory change in task-specific dystonia, and 3) if sequence tapping task captured similar level network changes as the dystonic writing task. Methods: We used the data-driven independent component analysis to compare functional network correlation with task time course and network connectivity during 3 motor tasks of writing, sequence tapping and finger flexion-extension. Results: Our analysis shows that in WC there is significantly decreased correlation of the basal ganglia and increased correlation of the orbitofrontal network to the writing task time course. Analysis of network connectivity showed significantly decreased connectivity of the basal ganglia to the left sensorimotor network specifically during the writing task in WC patients. We also showed significantly increased correlation of the cerebellum to the orbitofrontal and parietal networks. Conclusions: Collectively, our analysis supports a global network disorders in which the basal ganglia plays a primary role in the disease mechanism while the cerebellum plays a role as a compensatory network, partially through increased connectivity to the orbitofrontal network, a region important in motor planning. Our analysis also shows that the sequence tapping task does not capture the same network level dysfunction as the primary dystonic task. Our study is the first to apply a purely data-driven approach to understand network interactions in WC dystonia patients during task performance.