You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

 
Poster E46, Friday, November 10, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Interhemispheric functional connectivity predicts action semantic performance in stroke

Nicholas Riccardi1, Chris Rorden1,3, Julius Fridriksson2,3, Rutvik H. Desai1,3;1Department of Psychology, University of South Carolina, 2Department of Communication Science & Disorders, University of South Carolina, 3McCausland Center for Brain Imaging, University of South Carolina

Introduction: Behavioral and neuroimaging evidence from healthy participants largely supports embodied cognitive theories wherein sensorimotor systems participate in concept representation. This contrasts with amodal theories which argue that sensorimotor activity elicited by semantic tasks is epiphenomenal. Studies of patients with brain damage can play an important role in this debate by establishing a causal role of sensorimotor networks in language comprehension. Here, we examined the relationship between action verb comprehension and resting-state functional connectivity (RSFC) in stroke patients. With unilateral stroke, regions in the unaffected hemisphere sometimes play a compensatory role, mitigating the potential decline in cognitive performance. We hypothesized that impaired connectivity between the hemispheres, especially involving the motor cortex, would adversely affect action language comprehension. Methods: RSFC scans were collected from 48 chronic left-hemisphere (LH) stroke patients. Outside of the scanner, patients completed a verb semantic similarity judgment task (VSSJ) wherein 40 action (“to throw”) and 40 abstract (“to excuse”) verb triplets were presented. Participants were instructed to indicate via button press which of the bottom two verbs was most similar in meaning to the top. Accuracy difference scores were created for each patient by subtracting average accuracy in the abstract condition from that in the action condition. RSFC analysis was then conducted using a network of interest (NOI) consisting of seven bilateral regions: inferior frontal gyrus pars opercularis (IFGoper), IFG pars triangularis (IFGtri), supramarginal gyrus (SMG), precentral and postcentral gyri (PrC, PoC), and posterior middle and inferior temporal gyri (pMTG, pITG). For each patient, a correlation matrix of RSFC strength was generated for all left-to-left and left-to-right NOI regions. These correlation values were then used in a general linear model as a predictor of the VSSJ accuracy for action and abstract verbs separately with alpha set to 0.01, and for the difference scores with alpha set to 0.05. Significance was determined via permutation testing (5000 permutations) using NiiStat software. Results: Accuracy in the VSSJ task for action verbs was predicted by the strength of the following connections: 1) left IFGoper to right IFGoper, IFGtri, PrC, SMG; 2) left IFGtri to right IFGtri, IFGoper; 2) left PrC to right IFGoper, SMG; 3) left SMG to right SMG. Some of these results can reflect a general cognitive decline, affecting non-semantic factors. Examining the difference score revealed two interhemispheric RSFC links that were significant predictors of specifically action performance, such that lower connectivity predicted lower action relative to abstract verb performance: 1) left IFGoper to right PrC (p < .0005) and 2) left IFGoper to right SMG (p < .0005). Conclusions: PreC and SMG are strongly associated with action execution and planning. IFGoper, beyond a role in executive functions, is part of the mirror neuron system. Our results suggest that with damage to the LH, RH motor cortex participates in action verb comprehension, and when this compensatory ability is disrupted through impaired connectivity, a selective decline for action concepts is seen. These findings support a causal role of sensorimotor areas in concept representation, and highlight the role of network connectivity.

Topic Area: Speech Motor Control and Sensorimotor Integration

Back to Poster Schedule