Emir Turkes¹, Elisa Golfinopoulos, PhD², Frank H. Guenther, PhD³, Jason A. Tourville, PhD4
¹Boston University, ²Department of Speech, Language, & Hearing Sciences, Boston University, ³Department of Speech, Language, & Hearing Sciences, Boston University, 4Northeastern Department of Speech, Language, & Hearing Sciences, Center for Computational Neuroscience & Neural Technology, Boston University
Speech production is consistently associated with activity from a well-characterized network of brain regions. Functional interactions within that network, however, remain poorly understood. Here, resting state functional magnetic resonance imaging (fsFMRI) data from 497 neurotypical adults, collected as part of the Human Connectome Project (HCP), were used to assess intrinsic or resting state, functional connectivity (rsFC) between regions of interest (ROIs) within the speech production network. These findings were then compared to the inter-regional functional connectivity implicit in the DIVA neurocomputational model of speech production (Tourville & Guenther, 2011, Language and Cognitive Processes).
rsFC in the speech production network was derived from the minimally processed rsFMRI data in the HCP WU-Minn 500 Subjects released (http://humanconnectome.org/data/). The CONN toolbox (http://www.nitrc.org/projects/conn/) was used to further preprocess the data and for subsequent analyses. Cortical responses were mapped to a cortical surface reconstruction (FreeSurfer; http://freesurfer.net/). ROI boundaries were determined a priori based on a anatomical landmarks. The mean BOLD response time course within each seed ROI was then correlated with each surface vertex to generate surface rsFC maps for each ROI and subject. Second-level group comparisons were then performed to compare regional connectivity strength.
Preliminary findings indicate strong intrinsic functional coupling of the motor and sensory regions of the speech production network described in the DIVA model. Observed variation in regional functional connectivity within this network and strong coupling with regions outsides this network will guide further refinement of the model.
Keywords: Connectome, rsfMRI, Speech Network