Slide Slam P4
Cerebello-cerebral resting state functional connectivity in post-stroke aphasia
Joan Stilling1, Rachel Brenowitz1, Ji Hyun Kim1, Zafer Keser1, Donna C. Tippet1, Argye E. Hillis1, Rajani Sebastian1; 1JOHNS HOPKINS UNIVERSITY
INTRODUCTION: Resting state fMRI (rsfMRI) is an emerging method to understand network pathology in post-stroke aphasia (Klingbeil et al., 2019). Previous rsfMRI investigations indicate that disrupted functional connectivity across a distributed network of language and cognitive regions correlate with loss of language function post-stroke (e.g., Sebastian et al., 2016; Sandberg et al., 2017; Zhao et al., 2018). However, the influence of cerebello-cerebral resting state functional connectivity in aphasia recovery is poorly understood. The right cerebellum is a critical region involved in a broad range of language and cognitive functions (Marien et al., 2014). This study investigated the relationship between post-stroke naming deficits and cerebello-cerebral resting state functional connectivity. We hypothesized that weak resting state functional connectivity (low correlations) between the right cerebellum and the left hemisphere would be associated with poorer naming performance. METHODS: Twenty-six, right-handed participants (8 F, 18 M) with chronic (< 6 months post stroke) aphasia due to left hemisphere stroke were included in this study (mean age= 58.7 years). All participants underwent language testing including the short version of the Boston Naming Test (Mack, et al., 1992). Resting state scans were acquired on a Philips 3T scanner. Image preprocessing and statistical modeling were performed using the CONN toolbox implemented in MATLAB (Whitfield-Gabrieli & Nieto-Castanon, 2012). We conducted ROI-to-ROI analyses with a priori regions of interest (ROI) as the seed regions. The ROIs consisted of regions that comprise the language networks, including: right cerebellar Crus I, language seed ROI (Stoodley & Schmahmann, 2009); right cerebellar Crus II (Cereb2); bilateral cerebellar lobules seven (Cereb7) and eight (Cereb8); middle frontal gyrus (MidFG); inferior frontal pars opercularis (IFG oper) and pars triangularis (IFG tri); anterior and posterior superior temporal gyrus (STG), anterior and posterior middle temporal gyrus (MTG), anterior and posterior inferior temporal gyrus (ITG), angular gyrus (AG); anterior and posterior supramarginal gyrus (SMG); and lateral occipital cortex (LOC). Single-subject baseline functional connectivity parameters were extracted from CONN for each ROI that showed significant connectivity with the seed ROI and used as independent variables in a multiple linear regression model associating Boston Naming test score with baseline functional connectivity. RESULTS: Baseline functional connectivity analyses showed positive correlations between the right cerebellar Crus I and left cortical regions: ITG, SMG, LOC, and MTG. In multiple linear regression, controlling for time post stroke, age and lesion volume, baseline functional connectivity between right cerebellar Crus I and both the left anterior ITG (Beta = -22.52, p = 0.008) and right Cereb2 (Beta= 16.44, p= 0.021) were significant predictors of Boston Naming Test score. The overall model fit was R2 = 0.521 (p= 0.008). Left SMG, LOC and MTG were not significant predictors. CONCLUSION: These results suggest that cerebello-cerebral resting state functional connectivity may be an important factor in post-stroke aphasia recovery. As previous studies have mainly assessed cortical connectivity alone, this investigation proposes that functional connections between the right cerebellum and residual left hemisphere regions may also play a role in predicting naming ability post-stroke. Acknowledgements: NIH (NIDCD), R00 DC015554.